Will AI replace Actuaries, economists and statisticians?

Not as a job, but it is already doing parts of the work. Across the 234 official task statements scored for Actuaries, economists and statisticians (United Kingdom, SOC 2433), 62% of the importance-weighted core work is made of tasks today's AI could already do most of. The overall exposure score is 64 out of 100 (range 58–70, band: high). That is a statement about tasks, not about headcount: this measures what AI could do, not whether any employer adopts it, whether the law allows it, or whether doing the routine parts faster creates more demand for the human parts. Figures are from release 2026-q4.1.

Which tasks in “Actuaries, economists and statisticians” can AI already do?

The highest-scoring tasks in release 2026-q4.1 are: “Use programming languages and tools to manipulate and visualize data” (100/100, very high); “Write new functions or applications in programming languages to conduct analyses” (100/100, very high); “Implement data analysis algorithms” (100/100, very high). Each score comes from five published 0–4 ratings turned into a number by a published formula, and each carries the model's one-sentence reason on the page.

Which tasks in “Actuaries, economists and statisticians” stay human?

About 18% of this job's task weight sits in work that scores low for AI exposure. The lowest-scoring tasks in release 2026-q4.1 are: “Perform laboratory tests to produce reliable data for scientific investigations” (0/100, minimal); “Attend conferences to engage with the academic community” (3/100, minimal); “Act as an expert witness during legal proceedings” (7/100, minimal). Low scores usually mean the task needs a body in a room, a legally accountable human, or trust built in real time. Those are the three things the scoring rubric treats as gates rather than obstacles.

What should someone working in “Actuaries, economists and statisticians” do about AI?

Start from the ledger rather than the headline: 62% of this job's weighted core work is exposed, and roughly 18% is not. The practical move is to spend more of your week on the tasks that score low, the ones above, and to get fluent at directing AI through the tasks that score high, because those are the parts that change whether or not you are ready for them. This page does not predict your job, and nothing here is career advice tailored to you: the score describes the occupation, not the person.

How is the AI exposure score for Actuaries, economists and statisticians calculated?

Each official task statement for the occupation is rated on five published 0–4 dimensions (output replicability, physical embodiment, licensed accountability, real-time human trust, and data availability) by claude-opus-5 using scoring prompt task_scoring_v1.0. The model never writes the score; a published formula turns the five ratings into a 0–100 number, so every score can be recomputed by hand. The occupation figure is the importance-weighted mean across 234 scored tasks. The prompt, the rubric, the formula and the full dataset are published at https://futureproof.collab365.com/method and https://futureproof.collab365.com/data/2026-q4.1 under CC BY 4.0.

UK dataswitch to US

Actuaries, economists and statisticians

analysing data to support decision-making processes, advising fund managers on investment risks and payoffs for each recommendation and supervising economic research activities to analyse market trends. If that's your week, this page is about your job.

The honest answer

Most tasks in this job are the kind AI has learned to do: analysing data using mathematical and statistical methods to identify trends. The tasks, though, are not you.

Your move: three real directions from here ↓

Where the work sits, by task weight

shifting to AI: 62%
changing shape: 20%
staying human: 18%

These bars are tasks changing hands, not people being counted out. The ledger below shows which.

Whole-job exposure score 64 out of 100 (58–70 allowing for uncertainty): high exposure, across 234 scored tasks. The number is the support for the sentence above it, not a headline about anyone’s future.

How we know this

What is measured: Every published task statement for actuaries, economists and statisticians is rated on five dimensions: can a model produce the output, does the work need a body in a room, does it need a legally accountable person, does it depend on a person being trusted in the moment, and how much data exists. A published formula turns those five ratings into the score; the model never writes the number.

How the bar is built: Each task’s share of the bar is its published importance weight, so a task you do all day counts for more than one you do twice a year.

Release: 2026-q4.1, scores computed 2026-08-04. Read the full method.

It would be a lie to soften that; working collaboratively in teams and maintaining professional working relationships with colleagues and external stakeholders is what this work rebuilds around. The routes below start from it.

Your week, as this page understands it

Actuaries, economists and statisticians apply theoretical principles and practical techniques to assess risk and formulate probabilistic outcomes in order to inform economic and business policy, and to analyse and interpret data used to assist in the formulation of financial, business and economic policies in order to maximise growth or improve business performance. The job title says “actuaries”, “economists” or “statisticians”: officially one job, several names. The real job is the part underneath: working collaboratively in teams and maintaining professional working relationships with colleagues and external stakeholders. That is the thing someone has to be right about.

The exposed part of this job is specific, and we won’t pretend it is coming back. But actuaries, economists and statisticians is not one task. It is 234 scored ones, and they are nowhere near equally exposed. The most durable of them, on this evidence, is working collaboratively in teams and maintaining professional working relationships with colleagues and external stakeholders, and the ledger below shows exactly why.

Where the work sits, by task weight

shifting to AI: 62%
changing shape: 20%
staying human: 18%

These bars are tasks changing hands, not people being counted out. The ledger below shows which.

How we know this

How the bar is built: Each task’s share of the bar is its published importance weight, so a task you do all day counts for more than one you do twice a year.

Release: 2026-q4.1, scores computed 2026-08-04. Read the full method.

Your job, task by task

These are the official task statements for this occupation, in plain English, sorted by what the evidence says is happening to each one. The official wording sits under every line so you can check the rewrite against it.

One row is marked low confidence, so treat it as a ballpark rather than a fine measurement.
66 of this occupation's scored task statements carry a score that was measured under a different occupation's context, because the statement is shared between jobs and has only been scored once so far. Each one names the occupation it was measured under in the free routes below; none is presented as a measurement for this job.
Task clusters are not derived in this release, so the task-cluster field is empty and no Collab365 Space routing is attached to this occupation yet.

Shifting to AI

140 tasks

Tasks today’s tools can already do most of. This is the part we will not soften: where these rows are the bulk of your week, the week changes.

Analysing data using mathematical and statistical methods to identify trends
This is reading one thing and writing another: data in, a record out. That is the shape today's tools are built for.
importance 90 · 2433/04, 2433/05
Source: “Analyse data using mathematical and statistical methods to identify trends.” (UK task statement)
How this row was scored
Exposure score: 93 out of 100 (89–97 allowing for uncertainty): very high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Spotting trends with statistical methods is core computing work.
The five ratings: output a model can produce 4/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Analysing data to offer advanced insights and interpretations
This is reading one thing and writing another: data in, a record out. That is the shape today's tools are built for.
importance 80 · 2433/03, 2433/05
Source: “Analyse data to offer advanced insights and interpretations.” (UK task statement)
How this row was scored
Exposure score: 75 out of 100 (68–82 allowing for uncertainty): high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Software analyses data quickly and explains patterns, though an analyst checks the interpretation makes sense.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Identifying relationships and trends or any factors that could affect the results of research
This is reading one thing and writing another: relationships in, a record out. That is the shape today's tools are built for.
importance 85 · 2433/04, 2433/05
Source: “Identify relationships and trends or any factors that could affect the results of research.” (UK task statement)
How this row was scored
Exposure score: 75 out of 100 (68–82 allowing for uncertainty): high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Finding relationships and confounding factors in data is standard statistical work software performs well.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Manipulating large sets of data using statistical software to prepare for analysis
This is reading one thing and writing another: large sets of data in, a record out. That is the shape today's tools are built for.
importance 85 · 2433/04, 2433/05
Source: “Manipulate large sets of data using statistical software to prepare for analysis.” (UK task statement)
How this row was scored
Exposure score: 93 out of 100 (89–97 allowing for uncertainty): very high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Manipulating large datasets in statistical software is routine, scripted work.
The five ratings: output a model can produce 4/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.

Changing shape

48 tasks

Tasks where the machine takes the producing and a person keeps the checking, the deciding, or the answering-for-it. For most jobs this is the biggest group, and it is where "transformation, not termination" is literally visible.

Delivering oral or written presentations of the results of mathematical modelling and data analysis to management or other end users
The software now makes the first pass at oral, but the part that matters is a person saying it and standing behind it. So the job becomes checking and deciding rather than producing.
importance 80 · 2433/01, 2433/04
Source: “Deliver oral or written presentations of the results of mathematical modelling and data analysis to management or other end users.” (UK task statement)
How this row was scored
Exposure score: 53 out of 100 (46–60 allowing for uncertainty): partial exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; the value is that a specific person does it.
The rating behind it: Drafting the findings is easy for software, but presenting them to management is a live conversation.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 2/4 · how much data exists 3/4.
Presenting findings to clients to provide insights and recommendations
The software now makes the first pass at findings, but the part that matters is a person saying it and standing behind it. So the job becomes checking and deciding rather than producing.
importance 80 · 2433/01, 2433/05
Source: “Present findings to clients to provide insights and recommendations.” (UK task statement)
How this row was scored
Exposure score: 40 out of 100 (33–47 allowing for uncertainty): partial exposure, medium confidence.
Why it sits in this group: the same decision, made over and over; the value is that a specific person does it.
The rating behind it: Findings can be written up by software, but presenting them to a client is personal work.
The five ratings: output a model can produce 2/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 2/4 · how much data exists 3/4.
Producing tailored solutions to business problems using innovative methods
The software now makes the first pass at tailored solutions, but the deciding part still needs a person. So the job becomes checking and deciding rather than producing.
importance 80 · 2433/03, 2433/04
Source: “Produce tailored solutions to business problems using innovative methods.” (UK task statement)
How this row was scored
Exposure score: 58 out of 100 (51–65 allowing for uncertainty): partial exposure, medium confidence.
Why it sits in this group: the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Genuinely tailored, inventive solutions still lean on human insight, with tools offering starting points rather than finished answers.
The five ratings: output a model can produce 2/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.

Staying human

46 tasks

Tasks that stay with a person, because they happen in the physical world, because the rules need someone accountable, or because the value is that a specific person does them.

Presenting research findings at professional scientific conferences
This work happens in the physical world: research findings, in a real place. Software cannot follow it there.
importance 75 · 2433/03, 2433/05
Source: “Present research findings at professional scientific conferences.” (UK task statement)
How this row was scored
Exposure score: 14 out of 100 (7–21 allowing for uncertainty): minimal exposure, medium confidence.
Why it sits in this group: the same decision, made over and over; work that happens in the physical world; the value is that a specific person does it.
The rating behind it: Slides can be drafted automatically, but conference presenting and the questions afterwards are done by a person.
The five ratings: output a model can produce 1/4 · needs a body in a room 2/4 · needs an accountable person 0/4 · needs to be trusted in the moment 2/4 · how much data exists 3/4.
Working collaboratively in teams and maintaining professional working relationships with colleagues and external stakeholders
The value here is that a specific person handles teams and stands behind it. That is earned, not computed.
importance 70 · 2433/02, 2433/04
Source: “Work collaboratively in teams and maintain professional working relationships with colleagues and external stakeholders.” (UK task statement)
How this row was scored
Exposure score: 10 out of 100 (6–14 allowing for uncertainty): minimal exposure, high confidence.
Why it sits in this group: the value is that a specific person does it.
The rating behind it: Working relationships with colleagues and outside partners are built between people.
The five ratings: output a model can produce 1/4 · needs a body in a room 1/4 · needs an accountable person 0/4 · needs to be trusted in the moment 4/4 · how much data exists 2/4.
Presenting specific research findings to targeted audience groups
The value here is that a specific person handles specific research findings and stands behind it. That is earned, not computed.
importance 85 · 2433/03
Source: “Present specific research findings to targeted audience groups.” (UK task statement)
How this row was scored
Exposure score: 39 out of 100 (32–46 allowing for uncertainty): low exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; the value is that a specific person does it.
The rating behind it: Preparing the material is easy, but presenting findings to an audience is live work.
The five ratings: output a model can produce 3/4 · needs a body in a room 1/4 · needs an accountable person 0/4 · needs to be trusted in the moment 2/4 · how much data exists 3/4.

Show the other 224 tasks

Using programming languages and tools to manipulate and visualize data
shifting to AI
This is reading one thing and writing another: languages in, a record out. That is the shape today's tools are built for.
importance 85 · 2433/04
Source: “Use programming languages and tools to manipulate and visualize data.” (UK task statement)
How this row was scored
Exposure score: 100 out of 100 (96–100 allowing for uncertainty): very high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Writing code to manipulate and chart data is among the things software does best.
The five ratings: output a model can produce 4/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Writing new functions or applications in programming languages to conduct analyses
shifting to AI
This is reading one thing and writing another: new functions in, a record out. That is the shape today's tools are built for.
importance 85 · 2433/04
Source: “Write new functions or applications in programming languages to conduct analyses.” (UK task statement)
How this row was scored
Exposure score: 100 out of 100 (96–100 allowing for uncertainty): very high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Writing analysis code is one of the strongest and best-documented capabilities software has.
The five ratings: output a model can produce 4/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Implementing data analysis algorithms
shifting to AI
This is reading one thing and writing another: data analysis algorithms in, a record out. That is the shape today's tools are built for.
importance 85 · 2433/05
Source: “Implement data analysis algorithms.” (UK task statement)
How this row was scored
Exposure score: 100 out of 100 (96–100 allowing for uncertainty): very high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Implementing analysis algorithms in code is among software’s strongest and best-documented abilities.
The five ratings: output a model can produce 4/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Developing software programs for graphic data analysis
shifting to AI
This is reading one thing and writing another: software programs in, a record out. That is the shape today's tools are built for.
importance 70 · 2433/05
Source: “Develop software programs for graphic data analysis.” (UK task statement)
How this row was scored
Exposure score: 100 out of 100 (96–100 allowing for uncertainty): very high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Writing software for graphical data analysis is among the strongest things current tools do.
The five ratings: output a model can produce 4/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Analysing and evaluating data by performing complex calculations using computers
shifting to AI
This is reading one thing and writing another: data in, a record out. That is the shape today's tools are built for.
importance 90 · 2433/01
Source: “Analyse and evaluate data by performing complex calculations using computers.” (UK task statement)
How this row was scored
Exposure score: 93 out of 100 (89–97 allowing for uncertainty): very high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Complex calculations on data are what computers do best.
The five ratings: output a model can produce 4/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Analysing statistical data to calculate specific metrics
shifting to AI
This is reading one thing and writing another: statistical data in, a record out. That is the shape today's tools are built for.
importance 85 · 2433/01
Source: “Analyse statistical data to calculate specific metrics, such as accident rates for certain demographics.” (UK task statement)
How this row was scored
Exposure score: 93 out of 100 (89–97 allowing for uncertainty): very high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Calculating rates from statistical data is exactly the sort of arithmetic software does faster and more reliably.
The five ratings: output a model can produce 4/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Updating and managing internal and external documentation
shifting to AI
This is reading one thing and writing another: internal and external documentation in, a record out. That is the shape today's tools are built for.
importance 70 · 2433/01
Source: “Update and manage internal and external documentation.” (UK task statement)
How this row was scored
Exposure score: 93 out of 100 (86–100 allowing for uncertainty): very high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Keeping internal and external documentation current is routine updating software does accurately.
The five ratings: output a model can produce 4/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Sifting through collected data
shifting to AI
This is reading one thing and writing another: collected data in, a record out. That is the shape today's tools are built for.
importance 70 · 2433/02
Source: “Sift through collected data.” (UK task statement)
How this row was scored
Exposure score: 93 out of 100 (89–97 allowing for uncertainty): very high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Sifting through collected data is one of the things computers have always done faster than people.
The five ratings: output a model can produce 4/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Analysing large-scale data sets to identify patterns and insights
shifting to AI
This is reading one thing and writing another: large-scale data sets in, a record out. That is the shape today's tools are built for.
importance 85 · 2433/04
Source: “Analyse large-scale data sets to identify patterns and insights.” (UK task statement)
How this row was scored
Exposure score: 93 out of 100 (89–97 allowing for uncertainty): very high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Finding patterns in large datasets is one of the strongest things these tools do, working directly on workplace data.
The five ratings: output a model can produce 4/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Comparing models using statistical performance metrics
shifting to AI
This is reading one thing and writing another: models in, a record out. That is the shape today's tools are built for.
importance 85 · 2433/04
Source: “Compare models using statistical performance metrics, such as loss functions or proportion of explained variance.” (UK task statement)
How this row was scored
Exposure score: 93 out of 100 (89–97 allowing for uncertainty): very high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Comparing models on standard metrics is a calculation software runs automatically.
The five ratings: output a model can produce 4/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Applying feature selection algorithms to models predicting outcomes of interest
shifting to AI
This is reading one thing and writing another: feature selection algorithms in, a record out. That is the shape today's tools are built for.
importance 80 · 2433/04
Source: “Apply feature selection algorithms to models predicting outcomes of interest, such as sales, attrition, and healthcare use.” (UK task statement)
How this row was scored
Exposure score: 93 out of 100 (86–100 allowing for uncertainty): very high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Feature selection runs on published algorithms that modelling tools already apply automatically.
The five ratings: output a model can produce 4/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Identifying best practice in AI data systems
shifting to AI
This is reading one thing and writing another: best practice in, a record out. That is the shape today's tools are built for.
importance 80 · 2433/04
Source: “Identify best practice in AI data systems, data structures, and data architecture.” (UK task statement)
How this row was scored
Exposure score: 93 out of 100 (89–97 allowing for uncertainty): very high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Spotting established good practice in data systems is exactly what tools trained on technical writing do reliably.
The five ratings: output a model can produce 4/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Using specialist computer software to present spreadsheets and graphs
shifting to AI
This is reading one thing and writing another: specialist computer software in, a record out. That is the shape today's tools are built for.
importance 80 · 2433/04
Source: “Use specialist computer software to present spreadsheets and graphs.” (UK task statement)
How this row was scored
Exposure score: 93 out of 100 (89–97 allowing for uncertainty): very high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Building spreadsheets and charts from data is well-automated office work.
The five ratings: output a model can produce 4/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Analysing data to support decision-making processes
shifting to AI
This is reading one thing and writing another: data in, a record out. That is the shape today's tools are built for.
importance 95 · 2433/05
Source: “Analyse data to support decision-making processes.” (UK task statement)
How this row was scored
Exposure score: 93 out of 100 (89–97 allowing for uncertainty): very high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Analysing data to inform a decision is one of the clearest strengths of these tools.
The five ratings: output a model can produce 4/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Adjusting and weighting the raw data for processing
shifting to AI
This is reading one thing and writing another: the raw data in, a record out. That is the shape today's tools are built for.
importance 90 · 2433/05
Source: “Adjust and weight the raw data for processing.” (UK task statement)
How this row was scored
Exposure score: 93 out of 100 (89–97 allowing for uncertainty): very high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Weighting and adjusting raw data is a set calculation software performs exactly.
The five ratings: output a model can produce 4/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Creating data visualisations to report findings from medical research
shifting to AI
This is reading one thing and writing another: data visualisations in, a record out. That is the shape today's tools are built for.
importance 85 · 2433/05
Source: “Create data visualisations to report findings from medical research.” (UK task statement)
How this row was scored
Exposure score: 93 out of 100 (89–97 allowing for uncertainty): very high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Producing charts from research data is routine, scripted work.
The five ratings: output a model can produce 4/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Presenting data for review and analysis
shifting to AI
This is reading one thing and writing another: data in, a record out. That is the shape today's tools are built for.
importance 85 · 2433/05
Source: “Present data for review and analysis using mediums such as tables, charts, and graphs.” (UK task statement)
How this row was scored
Exposure score: 93 out of 100 (89–97 allowing for uncertainty): very high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Turning data into tables, charts and graphs is standard, automatic output.
The five ratings: output a model can produce 4/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Identifying and correcting data inaccuracies or mistakes
shifting to AI
This is reading one thing and writing another: data inaccuracies in, a record out. That is the shape today's tools are built for.
importance 80 · 2433/05
Source: “Identify and correct data inaccuracies or mistakes.” (UK task statement)
How this row was scored
Exposure score: 93 out of 100 (89–97 allowing for uncertainty): very high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Finding and correcting data errors is checking work software does faster and more consistently than people.
The five ratings: output a model can produce 4/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Organising information to prepare data for processing
shifting to AI
This is reading one thing and writing another: information in, a record out. That is the shape today's tools are built for.
importance 80 · 2433/05
Source: “Organise information to prepare data for processing.” (UK task statement)
How this row was scored
Exposure score: 93 out of 100 (89–97 allowing for uncertainty): very high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Organising information ready for processing is routine data handling software performs accurately.
The five ratings: output a model can produce 4/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Maintaining databases of biological data for research and analysis purposes
shifting to AI
This is reading one thing and writing another: databases of biological data in, a record out. That is the shape today's tools are built for.
importance 75 · 2433/05
Source: “Maintain databases of biological data for research and analysis purposes.” (UK task statement)
How this row was scored
Exposure score: 93 out of 100 (86–100 allowing for uncertainty): very high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Maintaining a research database is routine, systematic work software handles well.
The five ratings: output a model can produce 4/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Completing data analysis tasks using spreadsheets and specialist software
shifting to AI
This is reading one thing and writing another: data analysis tasks in, a record out. That is the shape today's tools are built for.
importance 85 · 2433/01
Source: “Complete data analysis tasks using spreadsheets and specialist software.” (UK task statement)
How this row was scored
Exposure score: 88 out of 100 (84–92 allowing for uncertainty): very high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Spreadsheet and software-based data analysis is now routinely done to a professional standard by machine, with review.
The five ratings: output a model can produce 4/4 · needs a body in a room 0/4 · needs an accountable person 1/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Preparing statistical data for reports to clients
shifting to AI
This is reading one thing and writing another: statistical data in, a record out. That is the shape today's tools are built for.
importance 85 · 2433/05
Source: “Prepare statistical data for reports to clients.” (UK task statement)
How this row was scored
Exposure score: 88 out of 100 (84–92 allowing for uncertainty): very high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Preparing statistics and tables for client reports is repeatable data work that software does reliably.
The five ratings: output a model can produce 4/4 · needs a body in a room 0/4 · needs an accountable person 1/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Calculating appropriate sample size in medical research to answer specific questions
shifting to AI
This is reading one thing and writing another: appropriate sample size in, a record out. That is the shape today's tools are built for.
importance 75 · 2433/05
Source: “Calculate appropriate sample size in medical research to answer specific questions.” (UK task statement)
How this row was scored
Exposure score: 88 out of 100 (81–95 allowing for uncertainty): very high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Sample size calculations are standard formulae, though medical research keeps a responsible statistician involved.
The five ratings: output a model can produce 4/4 · needs a body in a room 0/4 · needs an accountable person 1/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Explaining complex ideas and findings in a way that can be easily understood
shifting to AI
This is reading one thing and writing another: complex ideas in, a record out. That is the shape today's tools are built for.
importance 75 · 2433/01
Source: “Explain complex ideas and findings in a way that can be easily understood.” (UK task statement)
How this row was scored
Exposure score: 85 out of 100 (81–89 allowing for uncertainty): very high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Turning complicated findings into plain language is exactly the kind of writing software does reliably.
The five ratings: output a model can produce 4/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 1/4 · how much data exists 4/4.
Understanding and using complex mathematical formulae in advanced statistics and modelling
shifting to AI
This is reading one thing and writing another: complex mathematical formulae in, a record out. That is the shape today's tools are built for.
importance 85 · 2433/01
Source: “Understand and use complex mathematical formulae in advanced statistics and modelling.” (UK task statement)
How this row was scored
Exposure score: 83 out of 100 (79–87 allowing for uncertainty): very high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Working with advanced statistical formulae sits squarely in the territory where software is strongest.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Analysing national and international data using specialist software to identify economic trends
shifting to AI
This is reading one thing and writing another: national in, a record out. That is the shape today's tools are built for.
importance 90 · 2433/02
Source: “Analyse national and international data using specialist software to identify economic trends.” (UK task statement)
How this row was scored
Exposure score: 83 out of 100 (79–87 allowing for uncertainty): very high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: National and international economic data is public and abundant, and analysing it is a strong point for software.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Analysing problems using micro, macro and statistical techniques
shifting to AI
This is reading one thing and writing another: problems in, a record out. That is the shape today's tools are built for.
importance 85 · 2433/02
Source: “Analyse problems using micro, macro, and statistical techniques.” (UK task statement)
How this row was scored
Exposure score: 83 out of 100 (79–87 allowing for uncertainty): very high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Economic and statistical techniques are thoroughly documented and software applies them accurately.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Assessing past and presenting economic issues and trends
shifting to AI
This is reading one thing and writing another: past in, a record out. That is the shape today's tools are built for.
importance 85 · 2433/02
Source: “Assess past and present economic issues and trends.” (UK task statement)
How this row was scored
Exposure score: 83 out of 100 (79–87 allowing for uncertainty): very high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Assessing past and present economic trends draws on abundant published data and analysis.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Developing economic models to predict future economic outcomes
shifting to AI
This is reading one thing and writing another: economic models in, a record out. That is the shape today's tools are built for.
importance 85 · 2433/02
Source: “Develop economic models to predict future economic outcomes.” (UK task statement)
How this row was scored
Exposure score: 83 out of 100 (79–87 allowing for uncertainty): very high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Building forecasting models is mathematical work with plenty of published method behind it.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Producing rigorous economic analysis to inform decision-making processes using statistical and spreadsheet software
shifting to AI
This is reading one thing and writing another: rigorous economic analysis in, a record out. That is the shape today's tools are built for.
importance 85 · 2433/02
Source: “Produce rigorous economic analysis to inform decision-making processes using statistical and spreadsheet software.” (UK task statement)
How this row was scored
Exposure score: 83 out of 100 (79–87 allowing for uncertainty): very high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Statistical and spreadsheet analysis is exactly where software is strongest.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Writing technical reports on economic trends and forecasts
shifting to AI
This is reading one thing and writing another: technical reports in, a record out. That is the shape today's tools are built for.
importance 85 · 2433/02
Source: “Write technical reports on economic trends and forecasts.” (UK task statement)
How this row was scored
Exposure score: 83 out of 100 (79–87 allowing for uncertainty): very high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Technical reports on economic trends are written from public data, which software does quickly and well.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Developing forecasts using econometric and financial modelling techniques
shifting to AI
This is reading one thing and writing another: forecasts in, a record out. That is the shape today's tools are built for.
importance 80 · 2433/02
Source: “Develop forecasts using econometric and financial modelling techniques.” (UK task statement)
How this row was scored
Exposure score: 83 out of 100 (79–87 allowing for uncertainty): very high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Econometric and financial forecasting is well documented with abundant reference material.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Studying the socioeconomic impacts of new public policies
shifting to AI
This is reading one thing and writing another: the socioeconomic impacts of new public policies in, a record out. That is the shape today's tools are built for.
importance 80 · 2433/02
Source: “Study the socioeconomic impacts of new public policies, including proposed legislation, taxes, services, and regulations.” (UK task statement)
How this row was scored
Exposure score: 83 out of 100 (79–87 allowing for uncertainty): very high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Studying the effects of new legislation and taxes uses public documents and data in abundance.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Utilising various financial modelling techniques for economic analysis
shifting to AI
This is reading one thing and writing another: various financial modelling techniques in, a record out. That is the shape today's tools are built for.
importance 80 · 2433/02
Source: “Utilise various financial modelling techniques for economic analysis.” (UK task statement)
How this row was scored
Exposure score: 83 out of 100 (79–87 allowing for uncertainty): very high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Financial modelling for economic analysis is well documented with abundant reference material.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Conducting literature reviews to support research projects or academic studies
shifting to AI
This is reading one thing and writing another: literature reviews in, a record out. That is the shape today's tools are built for.
importance 75 · 2433/02
Source: “Conduct literature reviews to support research projects or academic studies.” (UK task statement)
How this row was scored
Exposure score: 83 out of 100 (76–90 allowing for uncertainty): very high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Literature reviewing is reading and summarising at scale, which software does quickly, though citations need checking.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Determining data collection methods for research projects or surveys
shifting to AI
This is reading one thing and writing another: data collection methods in, a record out. That is the shape today's tools are built for.
importance 75 · 2433/02, 2433/05
Source: “Determine data collection methods for research projects or surveys.” (UK task statement)
How this row was scored
Exposure score: 83 out of 100 (76–90 allowing for uncertainty): very high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Choosing a data collection method follows well-published research practice, which software advises on confidently.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Explaining research methodology to clarify the approach taken
shifting to AI
This is reading one thing and writing another: research methodology in, a record out. That is the shape today's tools are built for.
importance 75 · 2433/02
Source: “Explain research methodology to clarify the approach taken.” (UK task statement)
How this row was scored
Exposure score: 83 out of 100 (79–87 allowing for uncertainty): very high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Explaining a research method is well-documented writing software produces clearly.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Reviewing and critically evaluating published information relevant to the task
shifting to AI
This is reading one thing and writing another: this work in, a record out. That is the shape today's tools are built for.
importance 75 · 2433/02, 2433/05
Source: “Review and critically evaluate published information relevant to the task, identify evidence quality weaknesses, data gaps, and develop approaches to address these.” (UK task statement)
How this row was scored
Exposure score: 83 out of 100 (76–90 allowing for uncertainty): very high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Appraising published evidence and spotting gaps is careful reading, an area where software is strong.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Structuring, cleaning and manipulating data and designing and overseeing its collection and transformation
shifting to AI
This is reading one thing and writing another: data in, a record out. That is the shape today's tools are built for.
importance 75 · 2433/02
Source: “Structure, clean, and manipulate data, and design and oversee its collection and transformation.” (UK task statement)
How this row was scored
Exposure score: 83 out of 100 (79–87 allowing for uncertainty): very high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Structuring, cleaning and reshaping data is standard programming work software does very well.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Disseminating research findings through technical reports or scientific articles in journals
shifting to AI
This is reading one thing and writing another: research findings in, a record out. That is the shape today's tools are built for.
importance 70 · 2433/02
Source: “Disseminate research findings through technical reports or scientific articles in journals.” (UK task statement)
How this row was scored
Exposure score: 83 out of 100 (79–87 allowing for uncertainty): very high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Writing up findings for reports or journals is drafting work software does well from supplied results.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Analysing data to develop mathematical models
shifting to AI
This is reading one thing and writing another: data in, a record out. That is the shape today's tools are built for.
importance 85 · 2433/03
Source: “Analyse data to develop mathematical models.” (UK task statement)
How this row was scored
Exposure score: 83 out of 100 (79–87 allowing for uncertainty): very high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Building mathematical models from data is well-documented work software does strongly.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Assembling sets of mathematical assumptions for analysis
shifting to AI
This is reading one thing and writing another: sets of mathematical assumptions in, a record out. That is the shape today's tools are built for.
importance 85 · 2433/03
Source: “Assemble sets of mathematical assumptions for analysis.” (UK task statement)
How this row was scored
Exposure score: 83 out of 100 (76–90 allowing for uncertainty): very high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Setting out the assumptions behind an analysis is written, well-documented work software handles well.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Developing computational methods to solve problems in science
shifting to AI
This is reading one thing and writing another: computational methods in, a record out. That is the shape today's tools are built for.
importance 85 · 2433/03
Source: “Develop computational methods to solve problems in science, engineering, business, or industry.” (UK task statement)
How this row was scored
Exposure score: 83 out of 100 (79–87 allowing for uncertainty): very high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Developing computational methods is programming and mathematics, both areas of real strength for software.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Developing programmes using specialist mathematical software like Mathematica to perform mathematical functions
shifting to AI
This is reading one thing and writing another: programmes in, a record out. That is the shape today's tools are built for.
importance 85 · 2433/03
Source: “Develop programmes using specialist mathematical software like Mathematica to perform mathematical functions.” (UK task statement)
How this row was scored
Exposure score: 83 out of 100 (79–87 allowing for uncertainty): very high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Writing code for mathematical software is well documented and software produces it to a good standard.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Identifying solutions by learning and applying new methods
shifting to AI
This is reading one thing and writing another: solutions in, a record out. That is the shape today's tools are built for.
importance 85 · 2433/03
Source: “Identify solutions by learning and applying new methods, for example designing mathematical models that interpret data in a meaningful way.” (UK task statement)
How this row was scored
Exposure score: 83 out of 100 (76–90 allowing for uncertainty): very high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Designing models that make sense of data is documented mathematical work software does well.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Performing computations and applying methods of numerical analysis to data
shifting to AI
This is reading one thing and writing another: computations in, a record out. That is the shape today's tools are built for.
importance 85 · 2433/03
Source: “Perform computations and apply methods of numerical analysis to data.” (UK task statement)
How this row was scored
Exposure score: 83 out of 100 (79–87 allowing for uncertainty): very high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Numerical computation is exactly what software and AI do fastest and most reliably.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Researching mathematical information for analysis and problem-solving
shifting to AI
This is reading one thing and writing another: mathematical information in, a record out. That is the shape today's tools are built for.
importance 85 · 2433/03
Source: “Research mathematical information for analysis and problem-solving.” (UK task statement)
How this row was scored
Exposure score: 83 out of 100 (79–87 allowing for uncertainty): very high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Researching mathematical information is reading and retrieval, where software is fast and thorough.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Developing statistical models or simulations using statistical or modelling software
shifting to AI
This is reading one thing and writing another: statistical models in, a record out. That is the shape today's tools are built for.
importance 80 · 2433/03
Source: “Develop statistical models or simulations using statistical or modelling software.” (UK task statement)
How this row was scored
Exposure score: 83 out of 100 (79–87 allowing for uncertainty): very high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Statistical modelling in standard software is one of the things current tools do most reliably.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Keeping up to date with new mathematical developments
shifting to AI
This is reading one thing and writing another: to date in, a record out. That is the shape today's tools are built for.
importance 80 · 2433/03
Source: “Keep up to date with new mathematical developments.” (UK task statement)
How this row was scored
Exposure score: 83 out of 100 (79–87 allowing for uncertainty): very high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Keeping up with new mathematical work is reading and summarising published material.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Suggesting new ways to analyse data
shifting to AI
This is reading one thing and writing another: new ways in, a record out. That is the shape today's tools are built for.
importance 80 · 2433/03
Source: “Suggest new ways to analyse data.” (UK task statement)
How this row was scored
Exposure score: 83 out of 100 (76–90 allowing for uncertainty): very high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Suggesting new ways to analyse data draws on a huge published literature software knows well.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Exploring the consequences of each mathematical assumption set
shifting to AI
This is reading one thing and writing another: the consequences of each mathematical assumption set in, a record out. That is the shape today's tools are built for.
importance 75 · 2433/03
Source: “Explore the consequences of each mathematical assumption set.” (UK task statement)
How this row was scored
Exposure score: 83 out of 100 (76–90 allowing for uncertainty): very high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Working out what follows from a set of assumptions is exactly the kind of reasoning software does well.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Reading professional journals to enhance knowledge in the field
shifting to AI
This is reading one thing and writing another: professional journals in, a record out. That is the shape today's tools are built for.
importance 75 · 2433/03
Source: “Read professional journals to enhance knowledge in the field.” (UK task statement)
How this row was scored
Exposure score: 83 out of 100 (76–90 allowing for uncertainty): very high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Reading and digesting published journal articles is exactly what machines do quickly and thoroughly.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Developing, or applying dataing mining and machine learning algorithms
shifting to AI
This is reading one thing and writing another: machine learning algorithms in, a record out. That is the shape today's tools are built for.
importance 90 · 2433/04
Source: “Develop or apply data mining and machine learning algorithms.” (UK task statement)
How this row was scored
Exposure score: 83 out of 100 (76–90 allowing for uncertainty): very high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Applying known machine learning methods is well supported by tools, though genuinely new algorithm work still needs a researcher.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Developing reproducible analysis and robust code following software development standards
shifting to AI
This is reading one thing and writing another: reproducible analysis in, a record out. That is the shape today's tools are built for.
importance 85 · 2433/04
Source: “Develop reproducible analysis and robust code following software development standards.” (UK task statement)
How this row was scored
Exposure score: 83 out of 100 (79–87 allowing for uncertainty): very high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Writing reproducible, well-structured analysis code is a real strength of current software.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Reading scientific articles, conference papers or other sources of research to identify emerging analytic trends and technologies
shifting to AI
This is reading one thing and writing another: scientific articles in, a record out. That is the shape today's tools are built for.
importance 85 · 2433/04
Source: “Read scientific articles, conference papers, or other sources of research to identify emerging analytic trends and technologies.” (UK task statement)
How this row was scored
Exposure score: 83 out of 100 (76–90 allowing for uncertainty): very high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Research literature is public and plentiful, so software summarises it well, though judging what genuinely matters takes experience.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Reformulating models to improve prediction accuracy
shifting to AI
This is reading one thing and writing another: models in, a record out. That is the shape today's tools are built for.
importance 85 · 2433/04
Source: “Reformulate models to improve prediction accuracy as necessary.” (UK task statement)
How this row was scored
Exposure score: 83 out of 100 (76–90 allowing for uncertainty): very high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Retuning models to improve accuracy is well-documented and largely automated work.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Validating and testing models to ensure adequacy
shifting to AI
This is reading one thing and writing another: models in, a record out. That is the shape today's tools are built for.
importance 85 · 2433/04
Source: “Validate and test models to ensure adequacy.” (UK task statement)
How this row was scored
Exposure score: 83 out of 100 (76–90 allowing for uncertainty): very high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Model validation follows established procedures software runs reliably.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Designing and implementing appropriate data storage formats and associated database structure
shifting to AI
This is reading one thing and writing another: appropriate data storage formats in, a record out. That is the shape today's tools are built for.
importance 80 · 2433/04
Source: “Design and implement appropriate data storage formats and associated database structure.” (UK task statement)
How this row was scored
Exposure score: 83 out of 100 (79–87 allowing for uncertainty): very high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Database design is thoroughly documented technical work software produces to a good standard.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Sourcing and accessing diverse datasets for analysis using appropriate tools and techniques
shifting to AI
This is reading one thing and writing another: diverse datasets in, a record out. That is the shape today's tools are built for.
importance 80 · 2433/04
Source: “Source and access diverse datasets for analysis using appropriate tools and techniques.” (UK task statement)
How this row was scored
Exposure score: 83 out of 100 (76–90 allowing for uncertainty): very high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Finding and pulling in datasets is well-documented technical work software does efficiently.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Applying sampling techniques to determine groups to be surveyed or use complete enumeration methods
shifting to AI
This is reading one thing and writing another: techniques in, a record out. That is the shape today's tools are built for.
importance 75 · 2433/04, 2433/05
Source: “Apply sampling techniques to determine groups to be surveyed or use complete enumeration methods.” (UK task statement)
How this row was scored
Exposure score: 83 out of 100 (76–90 allowing for uncertainty): very high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Sampling methods are textbook statistics software applies accurately.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Curating biological data using suitable metadata
shifting to AI
This is reading one thing and writing another: biological data in, a record out. That is the shape today's tools are built for.
importance 70 · 2433/04
Source: “Curate biological data using suitable metadata, ontologies, and/or controlled vocabularies.” (UK task statement)
How this row was scored
Exposure score: 83 out of 100 (76–90 allowing for uncertainty): very high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Curating data with standard vocabularies and metadata is documented, rule-based work.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Designing opinion polls to collect data
shifting to AI
This is reading one thing and writing another: opinion polls in, a record out. That is the shape today's tools are built for.
importance 70 · 2433/04
Source: “Design opinion polls to collect data.” (UK task statement)
How this row was scored
Exposure score: 83 out of 100 (79–87 allowing for uncertainty): very high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Poll and survey design follows well-documented methods that software can apply to a stated research question.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Choosing appropriate computational infrastructure and database solutions -
shifting to AI
This is reading one thing and writing another: appropriate computational infrastructure in, a record out. That is the shape today's tools are built for.
importance 65 · 2433/04
Source: “Choose appropriate computational infrastructure and database solutions - including internal or external/cloud resources.” (UK task statement)
How this row was scored
Exposure score: 83 out of 100 (76–90 allowing for uncertainty): very high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Choosing cloud and database options is well-documented technical work with plenty of public guidance.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Determining whether statistical methods are appropriate based on user needs or researching questions
shifting to AI
This is reading one thing and writing another: this work in, a record out. That is the shape today's tools are built for.
importance 90 · 2433/05
Source: “Determine whether statistical methods are appropriate based on user needs or research questions.” (UK task statement)
How this row was scored
Exposure score: 83 out of 100 (76–90 allowing for uncertainty): very high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Choosing an appropriate statistical method is thoroughly documented in textbooks and the literature.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Using data analysis tools to process or produce reliable analysis
shifting to AI
This is reading one thing and writing another: data analysis tools in, a record out. That is the shape today's tools are built for.
importance 90 · 2433/05
Source: “Use data analysis tools to process or produce reliable analysis, interpretation, and evaluation of data.” (UK task statement)
How this row was scored
Exposure score: 83 out of 100 (79–87 allowing for uncertainty): very high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Processing and interpreting data with analysis tools is a core strength of current software.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Designing research projects using scientific techniques and historical data for efficient analyses
shifting to AI
This is reading one thing and writing another: research projects in, a record out. That is the shape today's tools are built for.
importance 85 · 2433/05
Source: “Design research projects using scientific techniques and historical data for efficient analyses.” (UK task statement)
How this row was scored
Exposure score: 83 out of 100 (76–90 allowing for uncertainty): very high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Designing a study from established techniques and historical data is well-documented work.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Developing algorithms based on statistical modelling procedures
shifting to AI
This is reading one thing and writing another: algorithms in, a record out. That is the shape today's tools are built for.
importance 85 · 2433/05
Source: “Develop algorithms based on statistical modelling procedures.” (UK task statement)
How this row was scored
Exposure score: 83 out of 100 (79–87 allowing for uncertainty): very high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Building statistical algorithms is well-documented, code-based work that machines produce to a high standard.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Adapting statistical methods to solve specific problems in economics
shifting to AI
This is reading one thing and writing another: statistical methods in, a record out. That is the shape today's tools are built for.
importance 80 · 2433/05
Source: “Adapt statistical methods to solve specific problems in economics.” (UK task statement)
How this row was scored
Exposure score: 83 out of 100 (76–90 allowing for uncertainty): very high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Adapting standard statistical methods to economic questions draws on a large published literature.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Designing experiments for research purposes
shifting to AI
This is reading one thing and writing another: experiments in, a record out. That is the shape today's tools are built for.
importance 80 · 2433/05
Source: “Design experiments for research purposes.” (UK task statement)
How this row was scored
Exposure score: 83 out of 100 (76–90 allowing for uncertainty): very high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Experimental design follows well-published principles software applies competently.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Preparing original research for presentation at conferences
shifting to AI
This is reading one thing and writing another: original research in, a record out. That is the shape today's tools are built for.
importance 75 · 2433/05
Source: “Prepare original research for presentation at conferences.” (UK task statement)
How this row was scored
Exposure score: 83 out of 100 (76–90 allowing for uncertainty): very high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Preparing research for a conference is writing and figure-making software does to a good standard.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Reviewing scientific literature to maintain current knowledge of clinical studies affairs
shifting to AI
This is reading one thing and writing another: scientific literature in, a record out. That is the shape today's tools are built for.
importance 75 · 2433/05
Source: “Review scientific literature to maintain current knowledge of clinical studies affairs.” (UK task statement)
How this row was scored
Exposure score: 83 out of 100 (79–87 allowing for uncertainty): very high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Reading and summarising published research is one of the things software does best.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Preparing data warehouses for storing data
shifting to AI
This is reading one thing and writing another: data warehouses in, a record out. That is the shape today's tools are built for.
importance 70 · 2433/05
Source: “Prepare data warehouses for storing data.” (UK task statement)
How this row was scored
Exposure score: 83 out of 100 (76–90 allowing for uncertainty): very high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Building data warehouses is well-documented technical work software produces competently.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Writing articles for professional journals and magazines
shifting to AI
This is reading one thing and writing another: articles in, a record out. That is the shape today's tools are built for.
importance 70 · 2433/05
Source: “Write articles for professional journals and magazines.” (UK task statement)
How this row was scored
Exposure score: 83 out of 100 (76–90 allowing for uncertainty): very high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Writing articles for journals and magazines is drafting work software does to a publishable-draft standard.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Preparing data for use in mathematical models and ensuring its quality and accuracy
shifting to AI
This is reading one thing and writing another: data in, a record out. That is the shape today's tools are built for.
importance 85 · 2433/01
Source: “Prepare data for use in mathematical models and ensure its quality and accuracy.” (UK task statement)
How this row was scored
Exposure score: 75 out of 100 (71–79 allowing for uncertainty): high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Preparing and checking model data is structured work software handles well using the firm’s own systems.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Analysing statistics from actuaries and other sources
shifting to AI
This is reading one thing and writing another: statistics in, a record out. That is the shape today's tools are built for.
importance 80 · 2433/01
Source: “Analyse statistics from actuaries and other sources.” (UK task statement)
How this row was scored
Exposure score: 75 out of 100 (71–79 allowing for uncertainty): high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Analysing statistics from several sources is data work software does quickly and reliably.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Identifying and accessing source data to inform mathematical models
shifting to AI
This is reading one thing and writing another: source data in, a record out. That is the shape today's tools are built for.
importance 80 · 2433/01
Source: “Identify and access source data to inform mathematical models.” (UK task statement)
How this row was scored
Exposure score: 75 out of 100 (71–79 allowing for uncertainty): high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Finding and pulling in source data for models is well-documented work software does efficiently.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Creating presentations to communicate findings to clients
shifting to AI
This is reading one thing and writing another: presentations in, a record out. That is the shape today's tools are built for.
importance 75 · 2433/01
Source: “Create presentations to communicate findings to clients.” (UK task statement)
How this row was scored
Exposure score: 75 out of 100 (71–79 allowing for uncertainty): high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Turning findings into a clear client deck is well within what software produces.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Writing reports for clients in a manner understandable to non-technical audiences
shifting to AI
This is reading one thing and writing another: reports in, a record out. That is the shape today's tools are built for.
importance 75 · 2433/01
Source: “Write reports for clients in a manner understandable to non-technical audiences.” (UK task statement)
How this row was scored
Exposure score: 75 out of 100 (71–79 allowing for uncertainty): high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Writing technical findings in plain language is one of the things software does most reliably.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Analysing consumer data to identify and forecast market trends
shifting to AI
This is reading one thing and writing another: consumer data in, a record out. That is the shape today's tools are built for.
importance 70 · 2433/01
Source: “Analyse consumer data to identify and forecast market trends.” (UK task statement)
How this row was scored
Exposure score: 75 out of 100 (71–79 allowing for uncertainty): high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Spotting and forecasting trends in consumer data is a standard analysis job for software.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Maintaining knowledge of current key trends in the actuarial profession to enhance advice and analysis capabilities
shifting to AI
This is reading one thing and writing another: knowledge of current key trends in, a record out. That is the shape today's tools are built for.
importance 70 · 2433/01
Source: “Maintain knowledge of current key trends in the actuarial profession to enhance advice and analysis capabilities.” (UK task statement)
How this row was scored
Exposure score: 75 out of 100 (68–82 allowing for uncertainty): high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Keeping up with professional developments is reading and summarising, which software does quickly.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Analysing data patterns to form insights and making policy recommendations
shifting to AI
This is reading one thing and writing another: data patterns in, a record out. That is the shape today's tools are built for.
importance 90 · 2433/02
Source: “Analyse data patterns to form insights and make policy recommendations.” (UK task statement)
How this row was scored
Exposure score: 75 out of 100 (71–79 allowing for uncertainty): high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Finding patterns in data and drawing out recommendations is core analysis work software does well.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Analysing data sets and trends to identify the comparative benefits and potential economic impact of investing in infrastructure projects
shifting to AI
This is reading one thing and writing another: data sets in, a record out. That is the shape today's tools are built for.
importance 85 · 2433/02
Source: “Analyse data sets and trends to identify the comparative benefits and potential economic impact of investing in infrastructure projects such as transport lines, hospitals, or schools.” (UK task statement)
How this row was scored
Exposure score: 75 out of 100 (68–82 allowing for uncertainty): high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Appraising the economic impact of infrastructure follows published methods and largely open datasets.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Applying economic theory and knowledge of market dynamics to assess the potential size and share of a market for a specific product or service
shifting to AI
This is reading one thing and writing another: economic theory in, a record out. That is the shape today's tools are built for.
importance 85 · 2433/02
Source: “Apply economic theory and knowledge of market dynamics to assess the potential size and share of a market for a specific product or service.” (UK task statement)
How this row was scored
Exposure score: 75 out of 100 (68–82 allowing for uncertainty): high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Market sizing follows well-known methods using data that is largely available.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Conducting research and analysis considering various factors
shifting to AI
This is reading one thing and writing another: research in, a record out. That is the shape today's tools are built for.
importance 85 · 2433/02
Source: “Conduct research and analysis considering various factors including political, environmental, and economic.” (UK task statement)
How this row was scored
Exposure score: 75 out of 100 (68–82 allowing for uncertainty): high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Research pulling together political, environmental and economic factors is reading and synthesis software does well.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Evaluating results to draw robust statistical and economic conclusions
shifting to AI
This is reading one thing and writing another: results in, a record out. That is the shape today's tools are built for.
importance 85 · 2433/02
Source: “Evaluate results to draw robust statistical and economic conclusions.” (UK task statement)
How this row was scored
Exposure score: 75 out of 100 (68–82 allowing for uncertainty): high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Drawing conclusions from statistical results follows documented reasoning software applies reliably.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Formulating recommendations to solve economic problems
shifting to AI
This is reading one thing and writing another: recommendations in, a record out. That is the shape today's tools are built for.
importance 85 · 2433/02
Source: “Formulate recommendations to solve economic problems.” (UK task statement)
How this row was scored
Exposure score: 75 out of 100 (68–82 allowing for uncertainty): high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Turning analysis into recommendations is written work software produces to a usable standard.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Preparing points of view for formulating economic policy
shifting to AI
This is reading one thing and writing another: points of view in, a record out. That is the shape today's tools are built for.
importance 85 · 2433/02
Source: “Prepare points of view for formulating economic policy.” (UK task statement)
How this row was scored
Exposure score: 75 out of 100 (68–82 allowing for uncertainty): high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Preparing a position for policy discussion is drafting work built on published evidence.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Developing economic guidelines and standards for forecasting trends
shifting to AI
This is reading one thing and writing another: economic guidelines in, a record out. That is the shape today's tools are built for.
importance 80 · 2433/02
Source: “Develop economic guidelines and standards for forecasting trends.” (UK task statement)
How this row was scored
Exposure score: 75 out of 100 (68–82 allowing for uncertainty): high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Forecasting guidelines and standards are documented material software can draft well.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Justifying conclusions drawn from research data to validate findings
shifting to AI
This is reading one thing and writing another: conclusions drawn in, a record out. That is the shape today's tools are built for.
importance 80 · 2433/02
Source: “Justify conclusions drawn from research data to validate findings.” (UK task statement)
How this row was scored
Exposure score: 75 out of 100 (68–82 allowing for uncertainty): high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Justifying conclusions from data follows documented reasoning software sets out clearly.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Reviewing and qualitying assure economic analysis carried out by others to ensure its robustness and accuracy
shifting to AI
This is reading one thing and writing another: assure economic analysis carried out in, a record out. That is the shape today's tools are built for.
importance 80 · 2433/02
Source: “Review and quality assure economic analysis carried out by others to ensure its robustness and accuracy.” (UK task statement)
How this row was scored
Exposure score: 75 out of 100 (68–82 allowing for uncertainty): high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Checking someone else’s analysis for errors and robustness is careful, rule-based review software does well.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Forecasting production and consumption of renewable resources and supply, consumption and depletion of non-renewable resources
shifting to AI
This is reading one thing and writing another: production in, a record out. That is the shape today's tools are built for.
importance 70 · 2433/02
Source: “Forecast production and consumption of renewable resources and supply, consumption, and depletion of non-renewable resources.” (UK task statement)
How this row was scored
Exposure score: 75 out of 100 (71–79 allowing for uncertainty): high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Forecasting resource supply and depletion is modeling work built on published data.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Writing social impact statements for natural resource policies
shifting to AI
This is reading one thing and writing another: social impact statements in, a record out. That is the shape today's tools are built for.
importance 60 · 2433/02
Source: “Write social impact statements for natural resource policies.” (UK task statement)
How this row was scored
Exposure score: 75 out of 100 (68–82 allowing for uncertainty): high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Social impact statements are structured documents written from published evidence.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Preparing reports detailing research or project outcomes
shifting to AI
This is reading one thing and writing another: reports detailing research in, a record out. That is the shape today's tools are built for.
importance 80 · 2433/03
Source: “Prepare reports detailing research or project outcomes.” (UK task statement)
How this row was scored
Exposure score: 75 out of 100 (68–82 allowing for uncertainty): high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Writing up research or project outcomes is structured drafting software does to a usable standard.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Developing and architecting robust data sourcing and processing systems for the organisation
shifting to AI
This is reading one thing and writing another: robust data in, a record out. That is the shape today's tools are built for.
importance 85 · 2433/04
Source: “Develop and architect robust data sourcing and processing systems for the organisation.” (UK task statement)
How this row was scored
Exposure score: 75 out of 100 (71–79 allowing for uncertainty): high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Data pipeline architecture follows well-known patterns tools reproduce, though the organisation's own sources still need human design decisions.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Conducting tests on experimental models to gather data
shifting to AI
This is reading one thing and writing another: tests in, a record out. That is the shape today's tools are built for.
importance 80 · 2433/04
Source: “Conduct tests on experimental models to gather data.” (UK task statement)
How this row was scored
Exposure score: 75 out of 100 (68–82 allowing for uncertainty): high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Running tests on models to gather data is scripted, repeatable work software handles well.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Criticallying evaluate and synthesise AI research findings to apply them within an organisational context
shifting to AI
This is reading one thing and writing another: evaluate in, a record out. That is the shape today's tools are built for.
importance 80 · 2433/04
Source: “Critically evaluate and synthesise AI research findings to apply them within an organisational context.” (UK task statement)
How this row was scored
Exposure score: 75 out of 100 (71–79 allowing for uncertainty): high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Reading and summarising research is a strong point for these tools; applying it to this organisation needs a specialist.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Designing technical roadmaps for data life-cycles ensuring appropriate support and business processes are in place
shifting to AI
This is reading one thing and writing another: technical roadmaps in, a record out. That is the shape today's tools are built for.
importance 80 · 2433/04
Source: “Design technical roadmaps for data life-cycles ensuring appropriate support and business processes are in place.” (UK task statement)
How this row was scored
Exposure score: 75 out of 100 (68–82 allowing for uncertainty): high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Roadmaps for handling data follow familiar patterns, so tools draft them well, though local processes need human checking.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Developing time and cost plans for large projects
shifting to AI
This is reading one thing and writing another: time in, a record out. That is the shape today's tools are built for.
importance 80 · 2433/04
Source: “Develop time and cost plans for large projects.” (UK task statement)
How this row was scored
Exposure score: 75 out of 100 (68–82 allowing for uncertainty): high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Project time and cost plans follow familiar patterns software drafts well from project details.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Identifying solutions to budgeting problems using data analysis
shifting to AI
This is reading one thing and writing another: solutions in, a record out. That is the shape today's tools are built for.
importance 80 · 2433/04
Source: “Identify solutions to budgeting problems using data analysis.” (UK task statement)
How this row was scored
Exposure score: 75 out of 100 (68–82 allowing for uncertainty): high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Analysing budget data to find savings is standard analysis work for software.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Optimising mechanisms for accessing and analysing complex datasets to deliver business outcomes
shifting to AI
This is reading one thing and writing another: mechanisms in, a record out. That is the shape today's tools are built for.
importance 80 · 2433/04
Source: “Optimise mechanisms for accessing and analysing complex datasets to deliver business outcomes.” (UK task statement)
How this row was scored
Exposure score: 75 out of 100 (71–79 allowing for uncertainty): high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Query and access tuning is well documented, so tools suggest strong improvements that an engineer confirms against the live system.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Recognizing and critically reviewing the format
shifting to AI
This is reading one thing and writing another: reviewing the format in, a record out. That is the shape today's tools are built for.
importance 80 · 2433/04
Source: “Recognize and critically review the format, scope, and limitations of different biological data.” (UK task statement)
How this row was scored
Exposure score: 75 out of 100 (68–82 allowing for uncertainty): high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Reviewing what different biological datasets can and cannot show is documented analytical work.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Designing scalable real-time analytical solutions for business problems using AI and related technologies
shifting to AI
This is reading one thing and writing another: scalable real-time analytical solutions in, a record out. That is the shape today's tools are built for.
importance 75 · 2433/04
Source: “Design scalable real-time analytical solutions for business problems using AI and related technologies.” (UK task statement)
How this row was scored
Exposure score: 75 out of 100 (71–79 allowing for uncertainty): high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Designing analytics systems is well covered by these tools, but a specialist tunes the design to the business's real data.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Identifying solutions to marketing decisions using data analysis
shifting to AI
This is reading one thing and writing another: solutions in, a record out. That is the shape today's tools are built for.
importance 75 · 2433/04
Source: “Identify solutions to marketing decisions using data analysis.” (UK task statement)
How this row was scored
Exposure score: 75 out of 100 (68–82 allowing for uncertainty): high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Using data to guide marketing decisions is standard analysis work software does well.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Producing a range of standard and non-standard statistical and data analysis reports in the Model Building phase
shifting to AI
This is reading one thing and writing another: a range of standard in, a record out. That is the shape today's tools are built for.
importance 75 · 2433/04
Source: “Produce a range of standard and non-standard statistical and data analysis reports in the Model Building phase.” (UK task statement)
How this row was scored
Exposure score: 75 out of 100 (68–82 allowing for uncertainty): high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Standard reports are easy to automate; the non-standard ones still need an analyst's shaping.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Analysing statistical data to identify significant differences among information sources
shifting to AI
This is reading one thing and writing another: statistical data in, a record out. That is the shape today's tools are built for.
importance 90 · 2433/05
Source: “Analyse statistical data to identify significant differences among information sources.” (UK task statement)
How this row was scored
Exposure score: 75 out of 100 (68–82 allowing for uncertainty): high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Testing for significant differences between sources is standard statistics software performs reliably.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Conducting research on consumer opinions and marketing strategies
shifting to AI
This is reading one thing and writing another: research in, a record out. That is the shape today's tools are built for.
importance 85 · 2433/05
Source: “Conduct research on consumer opinions and marketing strategies.” (UK task statement)
How this row was scored
Exposure score: 75 out of 100 (71–79 allowing for uncertainty): high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Gathering and summarising research on consumer opinion and strategy is desk work software does quickly.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Reporting results of statistical analyses in peer-reviewed papers and technical manuals
shifting to AI
This is reading one thing and writing another: results of statistical analyses in, a record out. That is the shape today's tools are built for.
importance 85 · 2433/05
Source: “Report results of statistical analyses in peer-reviewed papers and technical manuals.” (UK task statement)
How this row was scored
Exposure score: 75 out of 100 (68–82 allowing for uncertainty): high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Writing up analyses for papers and manuals is drafting work software does well, with the researcher shaping the argument.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Determining project plans and timelines for statistical aspects of biological research studies
shifting to AI
This is reading one thing and writing another: project plans in, a record out. That is the shape today's tools are built for.
importance 80 · 2433/05
Source: “Determine project plans and timelines for statistical aspects of biological research studies.” (UK task statement)
How this row was scored
Exposure score: 75 out of 100 (68–82 allowing for uncertainty): high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Planning the statistical stages of a study follows familiar patterns software drafts well.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Planning research studies in life sciences
shifting to AI
This is reading one thing and writing another: research studies in, a record out. That is the shape today's tools are built for.
importance 70 · 2433/05
Source: “Plan research studies in life sciences.” (UK task statement)
How this row was scored
Exposure score: 75 out of 100 (68–82 allowing for uncertainty): high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Planning a life-science study follows documented method using information held in the project.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Maintaining models for logistics uses
shifting to AI
This is reading one thing and writing another: models in, a record out. That is the shape today's tools are built for.
importance 60 · 2433/05
Source: “Maintain models for logistics uses such as demand forecasting.” (UK task statement)
How this row was scored
Exposure score: 75 out of 100 (68–82 allowing for uncertainty): high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Maintaining forecasting models is routine technical work using the organisation’s own data.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Conducting statistical analyses to quantify risk using statistical analysis software
shifting to AI
This is reading one thing and writing another: statistical analyses in, a record out. That is the shape today's tools are built for.
importance 90 · 2433/05
Source: “Conduct statistical analyses to quantify risk using statistical analysis software.” (UK task statement)
How this row was scored
Exposure score: 72 out of 100 (68–76 allowing for uncertainty): high exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Statistical modelling in standard software is well-documented work that machines now do quickly and accurately, subject to review.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 1/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Selecting, and applying statisticaling methods for medical research and interpreting the results
shifting to AI
This is reading one thing and writing another: methods in, a record out. That is the shape today's tools are built for.
importance 85 · 2433/05
Source: “Select and apply statistical methods for medical research and interpret the results.” (UK task statement)
How this row was scored
Exposure score: 72 out of 100 (65–79 allowing for uncertainty): high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Choosing and applying statistical methods is well documented, though medical research keeps a named statistician responsible.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 1/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Producing technical writing in medical research
shifting to AI
This is reading one thing and writing another: this work in, a record out. That is the shape today's tools are built for.
importance 70 · 2433/05
Source: “Produce technical writing in medical research.” (UK task statement)
How this row was scored
Exposure score: 72 out of 100 (65–79 allowing for uncertainty): high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Medical technical writing is highly automatable, though a named professional remains responsible for accuracy.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 1/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Keeping up to date with relevant academic literature
shifting to AI
This is reading one thing and writing another: to date in, a record out. That is the shape today's tools are built for.
importance 85 · 2433/02
Source: “Keep up to date with relevant academic literature, developments in data science, techniques (e.g. behavioural insights) and contribute to wider debates within the organisation and economic profession.” (UK task statement)
How this row was scored
Exposure score: 70 out of 100 (63–77 allowing for uncertainty): high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Keeping up with literature and new techniques is reading and summarising, though joining internal debates stays human.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 1/4 · how much data exists 4/4.
Assisting in designing scientific research studies to ensure robust data analysis
shifting to AI
This is reading one thing and writing another: designing scientific research studies in, a record out. That is the shape today's tools are built for.
importance 85 · 2433/05
Source: “Assist in designing scientific research studies to ensure robust data analysis.” (UK task statement)
How this row was scored
Exposure score: 70 out of 100 (63–77 allowing for uncertainty): high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Study design follows published principles, though it is agreed with the researchers involved.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 1/4 · how much data exists 4/4.
Constructing probability tables for events like fires
shifting to AI
This is reading one thing and writing another: probability tables in, a record out. That is the shape today's tools are built for.
importance 90 · 2433/01
Source: “Construct probability tables for events like fires, natural disasters, and unemployment, based on statistical data analysis.” (UK task statement)
How this row was scored
Exposure score: 66 out of 100 (59–73 allowing for uncertainty): high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Building probability tables from statistical data is exactly the kind of calculation software does quickly and accurately.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 1/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Assisting in pricing corporate security offerings for financial risk assessment
shifting to AI
This is reading one thing and writing another: pricing corporate security offerings in, a record out. That is the shape today's tools are built for.
importance 80 · 2433/01
Source: “Assist in pricing corporate security offerings for financial risk assessment.” (UK task statement)
How this row was scored
Exposure score: 66 out of 100 (59–73 allowing for uncertainty): high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Pricing security offerings follows documented methods, with the firm’s own data supplying the inputs.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 1/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Proposing solutions in engineering, the sciences and other fields using mathematical theories and techniques
shifting to AI
This is reading one thing and writing another: solutions in, a record out. That is the shape today's tools are built for.
importance 85 · 2433/03, 2433/04
Source: “Propose solutions in engineering, the sciences, and other fields using mathematical theories and techniques.” (UK task statement)
How this row was scored
Exposure score: 66 out of 100 (59–73 allowing for uncertainty): high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Mathematical modelling and proposed solutions can be drafted well, though a qualified specialist signs off technical recommendations.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 1/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Creating and handling datasets
shifting to AI
This is reading one thing and writing another: datasets in, a record out. That is the shape today's tools are built for.
importance 85 · 2433/04
Source: “Create and handle datasets for analysis, ensuring data quality and privacy.” (UK task statement)
How this row was scored
Exposure score: 66 out of 100 (59–73 allowing for uncertainty): high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Dataset preparation is routine technical work, though privacy rules keep an accountable person involved.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 1/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Handling, processing and analysing data ensuring integrity and adherence to commercial practices
shifting to AI
This is reading one thing and writing another: data ensuring integrity in, a record out. That is the shape today's tools are built for.
importance 85 · 2433/04
Source: “Handle, process, and analyse data ensuring integrity and adherence to commercial practices.” (UK task statement)
How this row was scored
Exposure score: 66 out of 100 (59–73 allowing for uncertainty): high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Processing and analysing data is highly automatable, with responsibility for handling it kept with a person.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 1/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Preparing data for submission to appropriate public bioinformatics data repositories
shifting to AI
This is reading one thing and writing another: data in, a record out. That is the shape today's tools are built for.
importance 80 · 2433/04
Source: “Prepare data for submission to appropriate public bioinformatics data repositories as required, being aware of IP and/or ethical and legal issues.” (UK task statement)
How this row was scored
Exposure score: 66 out of 100 (59–73 allowing for uncertainty): high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Repository submissions follow published formats, with a person responsible for the legal and ethical checks.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 1/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Interpreting trial data
shifting to AI
This is reading one thing and writing another: trial data in, a record out. That is the shape today's tools are built for.
importance 80 · 2433/05
Source: “Interpret trial data.” (UK task statement)
How this row was scored
Exposure score: 66 out of 100 (59–73 allowing for uncertainty): high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Interpreting trial data follows documented statistical methods and is checked before it goes anywhere.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 1/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Addressing the relationships of quantities
shifting to AI
It is the same call made over and over on the relationships of quantities, with a right answer to check it against. That is what a model is trained on.
importance 90 · 2433/03
Source: “Address the relationships of quantities, magnitudes, and forms through the use of numbers and symbols.” (UK task statement)
How this row was scored
Exposure score: 65 out of 100 (53–77 allowing for uncertainty): high exposure, low confidence.
Why it sits in this group: the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Symbolic work is thoroughly documented, though deciding which relationships matter is still the mathematician's call.
The five ratings: output a model can produce 2/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Designing encryption systems for secure information transmission
shifting to AI
It is the same call made over and over on encryption systems, with a right answer to check it against. That is what a model is trained on.
importance 80 · 2433/03
Source: “Design encryption systems for secure information transmission.” (UK task statement)
How this row was scored
Exposure score: 65 out of 100 (58–72 allowing for uncertainty): high exposure, medium confidence.
Why it sits in this group: the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Standard encryption designs are public, but designing a genuinely secure system needs deep expert judgement.
The five ratings: output a model can produce 2/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Using spreadsheets and models to present data and findings to stakeholders
shifting to AI
This is reading one thing and writing another: spreadsheets in, a record out. That is the shape today's tools are built for.
importance 85 · 2433/01
Source: “Use spreadsheets and models to present data and findings to stakeholders.” (UK task statement)
How this row was scored
Exposure score: 64 out of 100 (57–71 allowing for uncertainty): high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Building spreadsheets and charts is easy for software, though someone presents them to colleagues.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 1/4 · how much data exists 3/4.
Using data to create presentations and inform management decisions
shifting to AI
This is reading one thing and writing another: data in, a record out. That is the shape today's tools are built for.
importance 80 · 2433/01, 2433/02
Source: “Use data to create presentations and inform management decisions.” (UK task statement)
How this row was scored
Exposure score: 64 out of 100 (57–71 allowing for uncertainty): high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Turning data into presentations is a strong point for software, though someone puts them in front of managers.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 1/4 · how much data exists 3/4.
Responding to enquiries by providing relevant information to internal and external stakeholders
shifting to AI
This is reading one thing and writing another: enquiries in, a record out. That is the shape today's tools are built for.
importance 75 · 2433/01
Source: “Respond to enquiries by providing relevant information to internal and external stakeholders.” (UK task statement)
How this row was scored
Exposure score: 64 out of 100 (57–71 allowing for uncertainty): high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Answering enquiries with the right information suits software well, with a person handling the trickier ones.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 1/4 · how much data exists 3/4.
Advising on the suitability of alternative actions and resource allocation based on data analysis
shifting to AI
This is reading one thing and writing another: the suitability of alternative actions in, a record out. That is the shape today's tools are built for.
importance 85 · 2433/02
Source: “Advise on the suitability of alternative actions and resource allocation based on data analysis.” (UK task statement)
How this row was scored
Exposure score: 64 out of 100 (57–71 allowing for uncertainty): high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Advice on options can be drafted from the analysis, though decision-makers usually want to discuss it.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 1/4 · how much data exists 3/4.
Providing economic advice and analysis to inform the formulation of government policies
shifting to AI
This is reading one thing and writing another: economic advice in, a record out. That is the shape today's tools are built for.
importance 85 · 2433/02
Source: “Provide economic advice and analysis to inform the formulation of government policies related to budgeting, interest rates, and the economic impact of climate or technological change.” (UK task statement)
How this row was scored
Exposure score: 64 out of 100 (57–71 allowing for uncertainty): high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: The analysis is highly automatable, though government advice is delivered through conversations with officials.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 1/4 · how much data exists 3/4.
Developing strategies to communicate complex economic ideas to senior stakeholders
shifting to AI
This is reading one thing and writing another: strategies in, a record out. That is the shape today's tools are built for.
importance 80 · 2433/02
Source: “Develop strategies to communicate complex economic ideas to senior stakeholders.” (UK task statement)
How this row was scored
Exposure score: 64 out of 100 (57–71 allowing for uncertainty): high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Communication strategies can be drafted easily, though landing them with senior people is personal work.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 1/4 · how much data exists 3/4.
Effectivelying manage multiple projects to meet tight deadlines and maintain busy schedules
shifting to AI
This is reading one thing and writing another: manage multiple projects in, a record out. That is the shape today's tools are built for.
importance 65 · 2433/02
Source: “Effectively manage multiple projects to meet tight deadlines and maintain busy schedules.” (UK task statement)
How this row was scored
Exposure score: 64 out of 100 (57–71 allowing for uncertainty): high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Scheduling and tracking several projects is well-documented work that software organises efficiently.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 1/4 · how much data exists 3/4.
Presenting findings and insights from data analysis in a clear and impactful manner
shifting to AI
This is reading one thing and writing another: findings in, a record out. That is the shape today's tools are built for.
importance 80 · 2433/04
Source: “Present findings and insights from data analysis in a clear and impactful manner.” (UK task statement)
How this row was scored
Exposure score: 64 out of 100 (57–71 allowing for uncertainty): high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Charts and write-ups are easy to produce, though presenting insights well involves reading the audience.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 1/4 · how much data exists 3/4.
Identifying business issues and using data analysis to propose solutions for decision making
shifting to AI
This is reading one thing and writing another: business issues in, a record out. That is the shape today's tools are built for.
importance 75 · 2433/04
Source: “Identify business issues and use data analysis to propose solutions for decision making.” (UK task statement)
How this row was scored
Exposure score: 64 out of 100 (57–71 allowing for uncertainty): high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: The analysis is highly automatable, though identifying the business issue involves talking to people.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 1/4 · how much data exists 3/4.
Collaborativelying maintain technical standards in AI solutions for the organisation and its customers
shifting to AI
This is reading one thing and writing another: maintain technical standards in, a record out. That is the shape today's tools are built for.
importance 70 · 2433/04
Source: “Collaboratively maintain technical standards in AI solutions for the organisation and its customers.” (UK task statement)
How this row was scored
Exposure score: 64 out of 100 (57–71 allowing for uncertainty): high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Technical standards are well documented so drafting is easy for tools, but agreeing them with colleagues stays human.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 1/4 · how much data exists 3/4.
Designing the statistical framework for medical trials and researching projects
shifting to AI
This is reading one thing and writing another: the statistical framework in, a record out. That is the shape today's tools are built for.
importance 80 · 2433/05
Source: “Design the statistical framework for medical trials and research projects.” (UK task statement)
How this row was scored
Exposure score: 62 out of 100 (55–69 allowing for uncertainty): high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; someone qualified has to answer for it.
The rating behind it: Trial statistics follow documented standards, but a named statistician is accountable for the design.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 2/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Contributing to ongoing medical research publications
shifting to AI
This is reading one thing and writing another: ongoing medical research publications in, a record out. That is the shape today's tools are built for.
importance 75 · 2433/05
Source: “Contribute to ongoing medical research publications.” (UK task statement)
How this row was scored
Exposure score: 61 out of 100 (54–68 allowing for uncertainty): high exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Drafting is highly automatable, but journals require named human authors accountable for the work.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 1/4 · needs to be trusted in the moment 1/4 · how much data exists 4/4.
Explaining complex technical concepts in simple terms to both technical and non-technical audiences
changing shape
The software now makes the first pass at complex technical concepts, but the part that matters is a person saying it and standing behind it. So the job becomes checking and deciding rather than producing.
importance 80 · 2433/01
Source: “Explain complex technical concepts in simple terms to both technical and non-technical audiences.” (UK task statement)
How this row was scored
Exposure score: 58 out of 100 (51–65 allowing for uncertainty): partial exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; the value is that a specific person does it.
The rating behind it: Turning technical material into plain explanations is a particular strength of current software, though delivery is often face to face.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 2/4 · how much data exists 4/4.
Developing or adapting techniques for novel analytical questions
changing shape
The software now makes the first pass at techniques, but the deciding part still needs a person. So the job becomes checking and deciding rather than producing.
importance 75 · 2433/02
Source: “Develop or adapt techniques for novel analytical questions.” (UK task statement)
How this row was scored
Exposure score: 58 out of 100 (51–65 allowing for uncertainty): partial exposure, medium confidence.
Why it sits in this group: the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Adapting methods to genuinely new questions still needs an expert, though software gives useful starting points.
The five ratings: output a model can produce 2/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Improving existing analytical techniques for more accurate or extensive predictions
changing shape
The software now makes the first pass at analytical techniques, but the deciding part still needs a person. So the job becomes checking and deciding rather than producing.
importance 75 · 2433/02
Source: “Improve existing analytical techniques for more accurate or extensive predictions.” (UK task statement)
How this row was scored
Exposure score: 58 out of 100 (51–65 allowing for uncertainty): partial exposure, medium confidence.
Why it sits in this group: the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Improving analytical techniques needs expert insight, though software contributes useful suggestions.
The five ratings: output a model can produce 2/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Assisting in implementing quality assurance measures
changing shape
The software now makes the first pass at implementing quality assurance measures, but the deciding part still needs a person. So the job becomes checking and deciding rather than producing.
importance 70 · 2433/02
Source: “Assist in implementing quality assurance measures.” (UK task statement)
How this row was scored
Exposure score: 58 out of 100 (51–65 allowing for uncertainty): partial exposure, medium confidence.
Why it sits in this group: the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Quality assurance steps are documented, but applying them depends on how the organisation actually works.
The five ratings: output a model can produce 2/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Publishing scientific papers detailing research findings
changing shape
The software now makes the first pass at scientific papers detailing research findings, but the deciding part still needs a person. So the job becomes checking and deciding rather than producing.
importance 90 · 2433/03
Source: “Publish scientific papers detailing research findings.” (UK task statement)
How this row was scored
Exposure score: 58 out of 100 (51–65 allowing for uncertainty): partial exposure, medium confidence.
Why it sits in this group: the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Drafting and formatting a paper is well supported by software, but the new findings come from the researcher.
The five ratings: output a model can produce 2/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Writing grant proposals to obtain funding for research
changing shape
The software now makes the first pass at grant proposals, but the deciding part still needs a person. So the job becomes checking and deciding rather than producing.
importance 75 · 2433/03
Source: “Write grant proposals to obtain funding for research.” (UK task statement)
How this row was scored
Exposure score: 58 out of 100 (51–65 allowing for uncertainty): partial exposure, medium confidence.
Why it sits in this group: the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Grant writing follows well-known formats, so software gives a solid draft that the researcher must sharpen.
The five ratings: output a model can produce 2/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Assessing and quantifying risks, limitations and biases of AI applications within specific business contexts
changing shape
The software now makes the first pass at risks, limitations and biases of AI applications, but the deciding part still needs a person. So the job becomes checking and deciding rather than producing.
importance 70 · 2433/04
Source: “Assess and quantify risks, limitations, and biases of AI applications within specific business contexts.” (UK task statement)
How this row was scored
Exposure score: 58 out of 100 (51–65 allowing for uncertainty): partial exposure, medium confidence.
Why it sits in this group: the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: General guidance is widely published, but assessing risk in one business needs its specific context and data.
The five ratings: output a model can produce 2/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Writing research proposals for submission to external bodies
changing shape
The software now makes the first pass at research proposals, but the deciding part still needs a person. So the job becomes checking and deciding rather than producing.
importance 75 · 2433/05
Source: “Write research proposals for submission to external bodies.” (UK task statement)
How this row was scored
Exposure score: 58 out of 100 (51–65 allowing for uncertainty): partial exposure, medium confidence.
Why it sits in this group: the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Proposals can be drafted quickly, but a persuasive one turns on knowing the funder and the team's own work.
The five ratings: output a model can produce 2/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Ascertaining premium rates required to ensure payment of future benefits
changing shape
The software now makes the first pass at premium rates, but someone has to be answerable for the result, and it cannot be the software. So the job becomes checking and deciding rather than producing.
importance 90 · 2433/01
Source: “Ascertain premium rates required to ensure payment of future benefits.” (UK task statement)
How this row was scored
Exposure score: 56 out of 100 (49–63 allowing for uncertainty): partial exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; someone qualified has to answer for it.
The rating behind it: Premium calculations follow well-known methods, but a qualified actuary is expected to stand behind the rates.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 2/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Using mathematical modelling techniques and statistical concepts to determine probability and assess risks
changing shape
The software now makes the first pass at mathematical modelling techniques, but someone has to be answerable for the result, and it cannot be the software. So the job becomes checking and deciding rather than producing.
importance 90 · 2433/01
Source: “Use mathematical modelling techniques and statistical concepts to determine probability and assess risks, such as analysing pension scheme liabilities to price commercial insurance.” (UK task statement)
How this row was scored
Exposure score: 56 out of 100 (49–63 allowing for uncertainty): partial exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; someone qualified has to answer for it.
The rating behind it: The modelling itself is very automatable, though pricing and scheme liability work is owned by a qualified actuary.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 2/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Helping administer insurance, annuity and pension schemes, determining financial soundness and calculating premiums
changing shape
The software now makes the first pass at administer insurance, annuity and pension schemes, but someone has to be answerable for the result, and it cannot be the software. So the job becomes checking and deciding rather than producing.
importance 85 · 2433/01
Source: “Help administer insurance, annuity and pension schemes, determining financial soundness and calculating premiums.” (UK task statement)
How this row was scored
Exposure score: 56 out of 100 (49–63 allowing for uncertainty): partial exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; someone qualified has to answer for it.
The rating behind it: Scheme calculations are highly systematic, though a qualified actuary is required to stand behind the results.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 2/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Conducting valuations and preparing quarterly updates
changing shape
The software now makes the first pass at valuations, but someone has to be answerable for the result, and it cannot be the software. So the job becomes checking and deciding rather than producing.
importance 80 · 2433/01
Source: “Conduct valuations and prepare quarterly updates.” (UK task statement)
How this row was scored
Exposure score: 56 out of 100 (49–63 allowing for uncertainty): partial exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; someone qualified has to answer for it.
The rating behind it: Valuations run on set methods and timetables, but a qualified actuary is expected to own the numbers.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 2/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Determining policy contract provisions for each type of insurance
changing shape
The software now makes the first pass at policy contract provisions, but someone has to be answerable for the result, and it cannot be the software. So the job becomes checking and deciding rather than producing.
importance 75 · 2433/01
Source: “Determine policy contract provisions for each type of insurance.” (UK task statement)
How this row was scored
Exposure score: 56 out of 100 (49–63 allowing for uncertainty): partial exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; someone qualified has to answer for it.
The rating behind it: Contract provisions follow standard wordings, but insurers expect a qualified professional to settle them.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 2/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Monitoring risk within trading positions to prevent excessive risks in investment banking
changing shape
The software now makes the first pass at risk within trading positions, but someone has to be answerable for the result, and it cannot be the software. So the job becomes checking and deciding rather than producing.
importance 75 · 2433/01
Source: “Monitor risk within trading positions to prevent excessive risks in investment banking.” (UK task statement)
How this row was scored
Exposure score: 56 out of 100 (49–63 allowing for uncertainty): partial exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; someone qualified has to answer for it.
The rating behind it: Position risk monitoring is highly systematised, though regulated firms keep an accountable person in the loop.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 2/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Producing reports outlining findings and explaining risk positions
changing shape
The software now makes the first pass at reports outlining findings, but someone has to be answerable for the result, and it cannot be the software. So the job becomes checking and deciding rather than producing.
importance 75 · 2433/01
Source: “Produce reports outlining findings and explaining risk positions.” (UK task statement)
How this row was scored
Exposure score: 56 out of 100 (49–63 allowing for uncertainty): partial exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; someone qualified has to answer for it.
The rating behind it: The analysis and writing are automatable, but a qualified actuary must stand behind a stated risk position.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 2/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Collecting and recording data for analysis and researching purposes
changing shape
The software now makes the first pass at data, but the deciding part still needs a person. So the job becomes checking and deciding rather than producing.
importance 80 · 2433/02
Source: “Collect and record data for analysis and research purposes.” (UK task statement)
How this row was scored
Exposure score: 56 out of 100 (49–63 allowing for uncertainty): partial exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Collecting and recording data is routine, though some of it has to be gathered in person.
The five ratings: output a model can produce 3/4 · needs a body in a room 1/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Writing briefing notes
changing shape
The software now makes the first pass at notes, but someone has to be answerable for the result, and it cannot be the software. So the job becomes checking and deciding rather than producing.
importance 65 · 2433/02
Source: “Write briefing notes, technical or expert witness reports from investigation results for use as evidence in legal situations.” (UK task statement)
How this row was scored
Exposure score: 56 out of 100 (49–63 allowing for uncertainty): partial exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; someone qualified has to answer for it.
The rating behind it: The writing is very automatable, but an expert witness report must be authored by the named expert.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 2/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Presenting detailed reports and actionable recommendations to stakeholders for decision-making
changing shape
The software now makes the first pass at detailed reports, but the part that matters is a person saying it and standing behind it. So the job becomes checking and deciding rather than producing.
importance 85 · 2433/04
Source: “Present detailed reports and actionable recommendations to stakeholders for decision-making.” (UK task statement)
How this row was scored
Exposure score: 53 out of 100 (49–57 allowing for uncertainty): partial exposure, high confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; the value is that a specific person does it.
The rating behind it: Reports and recommendations are drafted well from the organisation’s data, though presenting them persuades people.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 2/4 · how much data exists 3/4.
Designing and conducting surveys to collect data
changing shape
The software now makes the first pass at surveys, but the deciding part still needs a person. So the job becomes checking and deciding rather than producing.
importance 75 · 2433/04
Source: “Design and conduct surveys to collect data.” (UK task statement)
How this row was scored
Exposure score: 53 out of 100 (46–60 allowing for uncertainty): partial exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Survey design is well documented and easily drafted, though fieldwork means reaching real respondents.
The five ratings: output a model can produce 3/4 · needs a body in a room 1/4 · needs an accountable person 0/4 · needs to be trusted in the moment 1/4 · how much data exists 4/4.
Enhancing awareness of the wider application of AI tools and technologies across the business so that opportunities
changing shape
The software now makes the first pass at awareness of the wider application of AI tools, but the part that matters is a person saying it and standing behind it. So the job becomes checking and deciding rather than producing.
importance 75 · 2433/04
Source: “Enhance awareness of the wider application of AI tools and technologies across the business so that opportunities for its use can be identified.” (UK task statement)
How this row was scored
Exposure score: 53 out of 100 (46–60 allowing for uncertainty): partial exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; the value is that a specific person does it.
The rating behind it: Explaining where new tools could help is partly a persuasion job, though the supporting material writes itself easily.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 2/4 · how much data exists 3/4.
Implementing digital technologies to achieve specific organisational objectives
changing shape
The software now makes the first pass at digital technologies, but the deciding part still needs a person. So the job becomes checking and deciding rather than producing.
importance 65 · 2433/01
Source: “Implement digital technologies to achieve specific organisational objectives.” (UK task statement)
How this row was scored
Exposure score: 49 out of 100 (42–56 allowing for uncertainty): partial exposure, medium confidence.
Why it sits in this group: the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Actually putting technology into an organisation involves coordinating people and systems, which tools assist rather than deliver.
The five ratings: output a model can produce 2/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 1/4 · how much data exists 3/4.
Analysing stakeholder needs using techniques based on stakeholder theory
changing shape
The software now makes the first pass at stakeholder needs, but the deciding part still needs a person. So the job becomes checking and deciding rather than producing.
importance 80 · 2433/02
Source: “Analyse stakeholder needs using techniques based on stakeholder theory.” (UK task statement)
How this row was scored
Exposure score: 49 out of 100 (42–56 allowing for uncertainty): partial exposure, medium confidence.
Why it sits in this group: the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Stakeholder analysis has documented methods, but what people actually want comes out in conversation.
The five ratings: output a model can produce 2/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 1/4 · how much data exists 3/4.
Anticipating analytical requirements through horizon scanning and proactively scope work programmes for clients and colleagues
changing shape
The software now makes the first pass at analytical requirements, but the deciding part still needs a person. So the job becomes checking and deciding rather than producing.
importance 75 · 2433/02
Source: “Anticipate analytical requirements through horizon scanning and proactively scope work programmes for clients and colleagues.” (UK task statement)
How this row was scored
Exposure score: 49 out of 100 (42–56 allowing for uncertainty): partial exposure, medium confidence.
Why it sits in this group: the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Horizon scanning is easy for software, but scoping work for clients depends on knowing them.
The five ratings: output a model can produce 2/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 1/4 · how much data exists 3/4.
Analysing client needs using a range of Operational Research methods
changing shape
The software now makes the first pass at client needs, but the deciding part still needs a person. So the job becomes checking and deciding rather than producing.
importance 85 · 2433/04
Source: “Analyse client needs using a range of Operational Research methods, adapting and developing them whilst understanding their limitations.” (UK task statement)
How this row was scored
Exposure score: 49 out of 100 (42–56 allowing for uncertainty): partial exposure, medium confidence.
Why it sits in this group: the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Adapting operational research methods to a client's problem, and knowing their limits, is expert judgement.
The five ratings: output a model can produce 2/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 1/4 · how much data exists 3/4.
Providing technical authority for the business regarding emerging opportunities for AI
changing shape
The software now makes the first pass at technical authority, but the deciding part still needs a person. So the job becomes checking and deciding rather than producing.
importance 80 · 2433/04
Source: “Provide technical authority for the business regarding emerging opportunities for AI.” (UK task statement)
How this row was scored
Exposure score: 49 out of 100 (42–56 allowing for uncertainty): partial exposure, medium confidence.
Why it sits in this group: the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Being the person the business trusts on new technology is a standing role, not an output a tool can hand over.
The five ratings: output a model can produce 2/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 1/4 · how much data exists 3/4.
Providing technical oversight for data warehousing technologies to meet business objectives
changing shape
The software now makes the first pass at technical oversight, but the deciding part still needs a person. So the job becomes checking and deciding rather than producing.
importance 75 · 2433/04
Source: “Provide technical oversight for data warehousing technologies to meet business objectives.” (UK task statement)
How this row was scored
Exposure score: 49 out of 100 (42–56 allowing for uncertainty): partial exposure, medium confidence.
Why it sits in this group: the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Technical oversight is a standing responsibility someone carries, even though tools can produce most of the underlying analysis.
The five ratings: output a model can produce 2/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 1/4 · how much data exists 3/4.
Implementing their own continued personal and professional development plan
changing shape
The software now makes the first pass at their own continued personal and professional development plan, but the deciding part still needs a person. So the job becomes checking and deciding rather than producing.
importance 70 · 2433/04
Source: “Implement their own continued personal and professional development plan.” (UK task statement)
How this row was scored
Exposure score: 49 out of 100 (42–56 allowing for uncertainty): partial exposure, medium confidence.
Why it sits in this group: the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: A personal development plan is about the individual's own choices and effort.
The five ratings: output a model can produce 2/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 1/4 · how much data exists 3/4.
Initiating new projects in an agile environment
changing shape
The software now makes the first pass at new projects, but the deciding part still needs a person. So the job becomes checking and deciding rather than producing.
importance 65 · 2433/04
Source: “Initiate new projects in an agile environment.” (UK task statement)
How this row was scored
Exposure score: 49 out of 100 (42–56 allowing for uncertainty): partial exposure, medium confidence.
Why it sits in this group: the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Starting a project involves getting people moving and committing resources, which tools support rather than perform.
The five ratings: output a model can produce 2/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 1/4 · how much data exists 3/4.
Carrying out research, often as part of a team
changing shape
The software now makes the first pass at research, often, but the deciding part still needs a person. So the job becomes checking and deciding rather than producing.
importance 80 · 2433/05
Source: “Carrying out research, often as part of a team.” (UK task statement)
How this row was scored
Exposure score: 49 out of 100 (42–56 allowing for uncertainty): partial exposure, medium confidence.
Why it sits in this group: the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Research is partly automatable, but working as part of a team involves colleagues throughout.
The five ratings: output a model can produce 2/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 1/4 · how much data exists 3/4.
Conducting original mathematics research
changing shape
The software now makes the first pass at original mathematics research, but the deciding part still needs a person. So the job becomes checking and deciding rather than producing.
importance 90 · 2433/03
Source: “Conduct original mathematics research.” (UK task statement)
How this row was scored
Exposure score: 48 out of 100 (41–55 allowing for uncertainty): partial exposure, medium confidence.
Why it sits in this group: the same decision, made over and over.
The rating behind it: Software helps with proofs and examples, but genuinely original mathematics still comes from the researcher.
The five ratings: output a model can produce 1/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Formulating novel mathematical principles to solve complex theoretical problems
changing shape
The software now makes the first pass at novel mathematical principles, but the deciding part still needs a person. So the job becomes checking and deciding rather than producing.
importance 90 · 2433/03
Source: “Formulate novel mathematical principles to solve complex theoretical problems.” (UK task statement)
How this row was scored
Exposure score: 48 out of 100 (41–55 allowing for uncertainty): partial exposure, medium confidence.
Why it sits in this group: the same decision, made over and over.
The rating behind it: Creating new mathematical principles is frontier work where software assists rather than delivers.
The five ratings: output a model can produce 1/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Examining theories of probability and inference to discover mathematical bases for new or improved methods of evaluating numerical data
changing shape
The software now makes the first pass at theories of probability, but the deciding part still needs a person. So the job becomes checking and deciding rather than producing.
importance 85 · 2433/05
Source: “Examine theories of probability and inference to discover mathematical bases for new or improved methods of evaluating numerical data.” (UK task statement)
How this row was scored
Exposure score: 48 out of 100 (41–55 allowing for uncertainty): partial exposure, medium confidence.
Why it sits in this group: the same decision, made over and over.
The rating behind it: Developing new mathematical foundations for statistics is frontier research where software assists rather than leads.
The five ratings: output a model can produce 1/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 4/4.
Determining cash reserves and liabilities necessary for future benefits
changing shape
The software now makes the first pass at cash reserves, but someone has to be answerable for the result, and it cannot be the software. So the job becomes checking and deciding rather than producing.
importance 85 · 2433/01
Source: “Determine cash reserves and liabilities necessary for future benefits.” (UK task statement)
How this row was scored
Exposure score: 47 out of 100 (40–54 allowing for uncertainty): partial exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; someone qualified has to answer for it.
The rating behind it: Reserving calculations are well defined, but the rules put a named actuary in charge of the figures.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 3/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Explaining changes in contract provisions to customers
changing shape
The software now makes the first pass at changes, but the part that matters is a person saying it and standing behind it. So the job becomes checking and deciding rather than producing.
importance 60 · 2433/01
Source: “Explain changes in contract provisions to customers.” (UK task statement)
How this row was scored
Exposure score: 46 out of 100 (39–53 allowing for uncertainty): partial exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; the value is that a specific person does it.
The rating behind it: Explaining a contract change is clear writing software handles well, though customers often want to talk it through.
The five ratings: output a model can produce 3/4 · needs a body in a room 0/4 · needs an accountable person 1/4 · needs to be trusted in the moment 2/4 · how much data exists 3/4.
Contextualising own and others' analyses by considering wider political
changing shape
The software now makes the first pass at own, but the deciding part still needs a person. So the job becomes checking and deciding rather than producing.
importance 80 · 2433/02
Source: “Contextualise own and others' analyses by considering wider political, social, economic, and environmental issues affecting clients and adapt to contextual changes.” (UK task statement)
How this row was scored
Exposure score: 43 out of 100 (36–50 allowing for uncertainty): partial exposure, medium confidence.
Why it sits in this group: mistakes that are cheap to catch.
The rating behind it: Setting analysis in its wider context depends on knowing the client and the politics around them.
The five ratings: output a model can produce 2/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 1/4 · how much data exists 2/4.
Assessing trade-offs between analysis quality
changing shape
The software now makes the first pass at trade-offs, but the deciding part still needs a person. So the job becomes checking and deciding rather than producing.
importance 75 · 2433/02
Source: “Assess trade-offs between analysis quality, impact, resources, and time.” (UK task statement)
How this row was scored
Exposure score: 43 out of 100 (36–50 allowing for uncertainty): partial exposure, medium confidence.
Why it sits in this group: mistakes that are cheap to catch.
The rating behind it: Trading off quality, time and resources depends on the organisation’s own priorities and pressures.
The five ratings: output a model can produce 2/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 1/4 · how much data exists 2/4.
Managing and maintaining quality levels in service delivery
changing shape
The software now makes the first pass at quality levels, but the deciding part still needs a person. So the job becomes checking and deciding rather than producing.
importance 70 · 2433/02
Source: “Manage and maintain quality levels in service delivery.” (UK task statement)
How this row was scored
Exposure score: 43 out of 100 (36–50 allowing for uncertainty): partial exposure, medium confidence.
Why it sits in this group: mistakes that are cheap to catch.
The rating behind it: Maintaining service quality depends on day-to-day judgement about people and delivery.
The five ratings: output a model can produce 2/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 1/4 · how much data exists 2/4.
Choosing approaches that maximise impact within resource constraints
changing shape
The software now makes the first pass at approaches, but the deciding part still needs a person. So the job becomes checking and deciding rather than producing.
importance 65 · 2433/02
Source: “Choose approaches that maximise impact within resource constraints.” (UK task statement)
How this row was scored
Exposure score: 43 out of 100 (36–50 allowing for uncertainty): partial exposure, medium confidence.
Why it sits in this group: mistakes that are cheap to catch.
The rating behind it: Choosing where to spend limited effort depends on the organisation’s priorities rather than published method.
The five ratings: output a model can produce 2/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 1/4 · how much data exists 2/4.
Identifying errors in service delivery and taking appropriate corrective action
changing shape
The software now makes the first pass at errors, but the deciding part still needs a person. So the job becomes checking and deciding rather than producing.
importance 65 · 2433/02
Source: “Identify errors in service delivery and take appropriate corrective action.” (UK task statement)
How this row was scored
Exposure score: 43 out of 100 (36–50 allowing for uncertainty): partial exposure, medium confidence.
Why it sits in this group: mistakes that are cheap to catch.
The rating behind it: Spotting service failures is partly automatable, but putting them right involves people and local judgement.
The five ratings: output a model can produce 2/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 1/4 · how much data exists 2/4.
Contributing to the development and ethical and legal conduct of AI systems and processes
changing shape
The software now makes the first pass at the development, but the deciding part still needs a person. So the job becomes checking and deciding rather than producing.
importance 75 · 2433/04
Source: “Contribute to the development and ethical and legal conduct of AI systems and processes, in line with organisational and regulatory requirements.” (UK task statement)
How this row was scored
Exposure score: 43 out of 100 (36–50 allowing for uncertainty): partial exposure, medium confidence.
Why it sits in this group: the same decision, made over and over; mistakes that are cheap to catch.
The rating behind it: Guidance and rules are published, but applying them to one organisation's systems needs internal knowledge and judgement.
The five ratings: output a model can produce 2/4 · needs a body in a room 0/4 · needs an accountable person 1/4 · needs to be trusted in the moment 1/4 · how much data exists 3/4.
Performing selection of endpoints/variables in medical research for specific disease areas
changing shape
The software now makes the first pass at selection of endpoints/variables, but someone has to be answerable for the result, and it cannot be the software. So the job becomes checking and deciding rather than producing.
importance 75 · 2433/05
Source: “Perform selection of endpoints/variables in medical research for specific disease areas.” (UK task statement)
How this row was scored
Exposure score: 43 out of 100 (36–50 allowing for uncertainty): partial exposure, medium confidence.
Why it sits in this group: the same decision, made over and over; someone qualified has to answer for it.
The rating behind it: Choosing endpoints needs clinical and disease-specific judgement from an accountable specialist.
The five ratings: output a model can produce 2/4 · needs a body in a room 0/4 · needs an accountable person 2/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Providing advice and consultation on economic relationships
changing shape
The software now makes the first pass at advice, but the part that matters is a person saying it and standing behind it. So the job becomes checking and deciding rather than producing.
importance 80 · 2433/02
Source: “Provide advice and consultation on economic relationships to businesses, public bodies, and other employers.” (UK task statement)
How this row was scored
Exposure score: 40 out of 100 (33–47 allowing for uncertainty): partial exposure, medium confidence.
Why it sits in this group: the same decision, made over and over; the value is that a specific person does it.
The rating behind it: The analysis is automatable, but consultation with employers rests on a working relationship.
The five ratings: output a model can produce 2/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 2/4 · how much data exists 3/4.
Practising continuous self-learning to keep up to date with industry trends and developments to enhance relevant skills and take responsibility for own professional development
changing shape
The software now makes the first pass at continuous, but the deciding part still needs a person. So the job becomes checking and deciding rather than producing.
importance 85 · 2433/04
Source: “Practise continuous self-learning to keep up to date with industry trends and developments to enhance relevant skills and take responsibility for own professional development.” (UK task statement)
How this row was scored
Exposure score: 40 out of 100 (33–47 allowing for uncertainty): partial exposure, medium confidence.
Why it sits in this group: the same decision, made over and over.
The rating behind it: Keeping your own skills current is something only the individual can do.
The five ratings: output a model can produce 1/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 3/4.
Managing operational research projects
changing shape
The software now makes the first pass at operational research projects, but the part that matters is a person saying it and standing behind it. So the job becomes checking and deciding rather than producing.
importance 80 · 2433/04
Source: “Manage operational research projects.” (UK task statement)
How this row was scored
Exposure score: 40 out of 100 (33–47 allowing for uncertainty): partial exposure, medium confidence.
Why it sits in this group: the same decision, made over and over; the value is that a specific person does it.
The rating behind it: Running a project means coordinating people and decisions, not producing a single document.
The five ratings: output a model can produce 2/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 2/4 · how much data exists 3/4.
Collaborating with multi-disciplinary colleagues to design life-science experiments that will generate data suitable for subsequent bioinformatics analysis
changing shape
The software now makes the first pass at multi-disciplinary colleagues, but the part that matters is a person saying it and standing behind it. So the job becomes checking and deciding rather than producing.
importance 75 · 2433/04
Source: “Collaborate with multi-disciplinary colleagues to design life-science experiments that will generate data suitable for subsequent bioinformatics analysis.” (UK task statement)
How this row was scored
Exposure score: 40 out of 100 (33–47 allowing for uncertainty): partial exposure, medium confidence.
Why it sits in this group: the same decision, made over and over; the value is that a specific person does it.
The rating behind it: Experimental design is documented, but agreeing it happens across a table with laboratory colleagues.
The five ratings: output a model can produce 2/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 2/4 · how much data exists 3/4.
Explaining economic impact of policies to the public
staying human
The value here is that a specific person handles economic impact of policies and stands behind it. That is earned, not computed.
importance 80 · 2433/02
Source: “Explain economic impact of policies to the public.” (UK task statement)
How this row was scored
Exposure score: 39 out of 100 (32–46 allowing for uncertainty): low exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; the value is that a specific person does it.
The rating behind it: Clear explanations can be drafted, but the public listens to a person they find credible.
The five ratings: output a model can produce 3/4 · needs a body in a room 1/4 · needs an accountable person 0/4 · needs to be trusted in the moment 2/4 · how much data exists 3/4.
Delivering oral and visual presentations tailored for non-economist decision-makers
staying human
The value here is that a specific person handles oral and stands behind it. That is earned, not computed.
importance 70 · 2433/02
Source: “Deliver oral and visual presentations tailored for non-economist decision-makers.” (UK task statement)
How this row was scored
Exposure score: 39 out of 100 (32–46 allowing for uncertainty): low exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; the value is that a specific person does it.
The rating behind it: Slides can be drafted automatically, but delivering them to decision-makers is a live, personal job.
The five ratings: output a model can produce 3/4 · needs a body in a room 1/4 · needs an accountable person 0/4 · needs to be trusted in the moment 2/4 · how much data exists 3/4.
Presenting research findings to relevant stakeholders to support project development
staying human
The value here is that a specific person handles research findings and stands behind it. That is earned, not computed.
importance 80 · 2433/03
Source: “Present research findings to relevant stakeholders to support project development.” (UK task statement)
How this row was scored
Exposure score: 39 out of 100 (32–46 allowing for uncertainty): low exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; the value is that a specific person does it.
The rating behind it: The slides can be prepared automatically, but presenting to stakeholders and answering them is live.
The five ratings: output a model can produce 3/4 · needs a body in a room 1/4 · needs an accountable person 0/4 · needs to be trusted in the moment 2/4 · how much data exists 3/4.
Delivering presentations to senior management on these findings
staying human
The value here is that a specific person handles presentations and stands behind it. That is earned, not computed.
importance 75 · 2433/03
Source: “Deliver presentations to senior management on these findings.” (UK task statement)
How this row was scored
Exposure score: 39 out of 100 (32–46 allowing for uncertainty): low exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; the value is that a specific person does it.
The rating behind it: Slides and findings write themselves from the data; standing in front of directors and answering their questions does not.
The five ratings: output a model can produce 3/4 · needs a body in a room 1/4 · needs an accountable person 0/4 · needs to be trusted in the moment 2/4 · how much data exists 3/4.
Presenting statistical and non-statistical results using visual aids
staying human
The value here is that a specific person handles statistical and stands behind it. That is earned, not computed.
importance 85 · 2433/05
Source: “Present statistical and non-statistical results using visual aids to clients, colleagues, and students at meetings or conferences.” (UK task statement)
How this row was scored
Exposure score: 39 out of 100 (32–46 allowing for uncertainty): low exposure, medium confidence.
Why it sits in this group: reading one thing and writing another; the same decision, made over and over; the value is that a specific person does it.
The rating behind it: Producing the visuals is easy, but presenting to clients and colleagues is live work.
The five ratings: output a model can produce 3/4 · needs a body in a room 1/4 · needs an accountable person 0/4 · needs to be trusted in the moment 2/4 · how much data exists 3/4.
Advising on benefit and remuneration policies
staying human
The value here is that a specific person handles benefit and stands behind it. That is earned, not computed.
importance 80 · 2433/01
Source: “Advise on benefit and remuneration policies.” (UK task statement)
How this row was scored
Exposure score: 35 out of 100 (28–42 allowing for uncertainty): low exposure, medium confidence.
Why it sits in this group: the same decision, made over and over; the value is that a specific person does it.
The rating behind it: Benchmarks and rules are documented, but tailoring reward policy to one employer involves negotiation and judgement.
The five ratings: output a model can produce 2/4 · needs a body in a room 0/4 · needs an accountable person 1/4 · needs to be trusted in the moment 2/4 · how much data exists 3/4.
Assisting businesses in implementing strategic plans
staying human
The value here is that a specific person handles businesses and stands behind it. That is earned, not computed.
importance 75 · 2433/01
Source: “Assist businesses in implementing strategic plans.” (UK task statement)
How this row was scored
Exposure score: 35 out of 100 (28–42 allowing for uncertainty): low exposure, medium confidence.
Why it sits in this group: the value is that a specific person does it.
The rating behind it: Plans can be drafted, but helping a business actually adopt them depends on working alongside people.
The five ratings: output a model can produce 2/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 2/4 · how much data exists 2/4.
Helping improve existing business in collaboration with accounts
staying human
The value here is that a specific person handles improve existing business and stands behind it. That is earned, not computed.
importance 75 · 2433/01
Source: “Help improve existing business in collaboration with accounts, claims experts, and senior management.” (UK task statement)
How this row was scored
Exposure score: 35 out of 100 (28–42 allowing for uncertainty): low exposure, medium confidence.
Why it sits in this group: the value is that a specific person does it.
The rating behind it: Improving the business happens through discussion with claims, accounts and senior colleagues who hold the context.
The five ratings: output a model can produce 2/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 2/4 · how much data exists 2/4.
Providing technical support to a team of actuaries
staying human
The value here is that a specific person handles technical support and stands behind it. That is earned, not computed.
importance 70 · 2433/01
Source: “Provide technical support to a team of actuaries.” (UK task statement)
How this row was scored
Exposure score: 35 out of 100 (28–42 allowing for uncertainty): low exposure, medium confidence.
Why it sits in this group: the same decision, made over and over; the value is that a specific person does it.
The rating behind it: Technical support to colleagues depends on knowing their models and working alongside them.
The five ratings: output a model can produce 2/4 · needs a body in a room 0/4 · needs an accountable person 1/4 · needs to be trusted in the moment 2/4 · how much data exists 3/4.
Developing and implementing economic policies
staying human
The value here is that a specific person handles economic policies and stands behind it. That is earned, not computed.
importance 85 · 2433/02
Source: “Develop and implement economic policies.” (UK task statement)
How this row was scored
Exposure score: 35 out of 100 (28–42 allowing for uncertainty): low exposure, medium confidence.
Why it sits in this group: the value is that a specific person does it.
The rating behind it: Drafting policy is well within reach of software, but putting it into practice is an institutional job.
The five ratings: output a model can produce 2/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 2/4 · how much data exists 2/4.
Collaborating with colleagues and stakeholders to identify measurement needs and make informed decisions
staying human
The value here is that a specific person handles colleagues and stands behind it. That is earned, not computed.
importance 80 · 2433/04
Source: “Collaborate with colleagues and stakeholders to identify measurement needs and make informed decisions.” (UK task statement)
How this row was scored
Exposure score: 35 out of 100 (28–42 allowing for uncertainty): low exposure, medium confidence.
Why it sits in this group: the value is that a specific person does it.
The rating behind it: Agreeing what to measure happens through discussion with colleagues who hold the business context.
The five ratings: output a model can produce 2/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 2/4 · how much data exists 2/4.
Identifying problems in collaboration with senior managers and decision makers
staying human
The value here is that a specific person handles problems and stands behind it. That is earned, not computed.
importance 80 · 2433/04
Source: “Identify problems in collaboration with senior managers and decision makers.” (UK task statement)
How this row was scored
Exposure score: 35 out of 100 (28–42 allowing for uncertainty): low exposure, medium confidence.
Why it sits in this group: the value is that a specific person does it.
The rating behind it: Working out the real problem with senior managers depends on conversations with them.
The five ratings: output a model can produce 2/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 2/4 · how much data exists 2/4.
Planning, organizing and managing resources effectively to successfully complete data science projects
staying human
The value here is that a specific person handles resources and stands behind it. That is earned, not computed.
importance 80 · 2433/04
Source: “Plan, organize, and manage resources effectively to successfully complete data science projects.” (UK task statement)
How this row was scored
Exposure score: 35 out of 100 (28–42 allowing for uncertainty): low exposure, medium confidence.
Why it sits in this group: the value is that a specific person does it.
The rating behind it: Running a project is about coordinating people and priorities rather than producing a document.
The five ratings: output a model can produce 2/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 2/4 · how much data exists 2/4.
Consulting with clients to agree on data collection methods
staying human
The value here is that a specific person handles clients and stands behind it. That is earned, not computed.
importance 80 · 2433/05
Source: “Consult with clients to agree on data collection methods, taking into account ethical and legislative considerations.” (UK task statement)
How this row was scored
Exposure score: 35 out of 100 (28–42 allowing for uncertainty): low exposure, medium confidence.
Why it sits in this group: the same decision, made over and over; the value is that a specific person does it.
The rating behind it: Agreeing collection methods with clients is a conversation about their needs and legal duties.
The five ratings: output a model can produce 2/4 · needs a body in a room 0/4 · needs an accountable person 1/4 · needs to be trusted in the moment 2/4 · how much data exists 3/4.
Assigning work to biostatistical assistants or programmers
staying human
The value here is that a specific person handles work and stands behind it. That is earned, not computed.
importance 70 · 2433/05
Source: “Assign work to biostatistical assistants or programmers.” (UK task statement)
How this row was scored
Exposure score: 35 out of 100 (28–42 allowing for uncertainty): low exposure, medium confidence.
Why it sits in this group: the value is that a specific person does it.
The rating behind it: Assigning work to a team is about managing people, not producing a document.
The five ratings: output a model can produce 2/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 2/4 · how much data exists 2/4.
Teaching postgraduate or continuing education courses or seminars in biostatistics
staying human
The value here is that a specific person handles postgraduate and stands behind it. That is earned, not computed.
importance 60 · 2433/05
Source: “Teach postgraduate or continuing education courses or seminars in biostatistics.” (UK task statement)
How this row was scored
Exposure score: 34 out of 100 (27–41 allowing for uncertainty): low exposure, medium confidence.
Why it sits in this group: the same decision, made over and over; the value is that a specific person does it.
The rating behind it: Materials are easy to prepare, but teaching a postgraduate course is live work with students.
The five ratings: output a model can produce 2/4 · needs a body in a room 1/4 · needs an accountable person 0/4 · needs to be trusted in the moment 2/4 · how much data exists 4/4.
Determining or helping determine company policy in actuarial contexts
staying human
The rules require a named, qualified person to answer for determine company policy, and that person cannot be a piece of software.
importance 80 · 2433/01
Source: “Determine or help determine company policy in actuarial contexts.” (UK task statement)
How this row was scored
Exposure score: 32 out of 100 (25–39 allowing for uncertainty): low exposure, medium confidence.
Why it sits in this group: someone qualified has to answer for it.
The rating behind it: Policy setting depends on the firm’s own priorities and needs an accountable professional in the room.
The five ratings: output a model can produce 2/4 · needs a body in a room 0/4 · needs an accountable person 2/4 · needs to be trusted in the moment 1/4 · how much data exists 2/4.
Developing new financial products
staying human
The rules require a named, qualified person to answer for new financial products, and that person cannot be a piece of software.
importance 70 · 2433/01
Source: “Develop new financial products.” (UK task statement)
How this row was scored
Exposure score: 32 out of 100 (25–39 allowing for uncertainty): low exposure, medium confidence.
Why it sits in this group: someone qualified has to answer for it.
The rating behind it: New products need commercial judgement and a qualified professional behind the pricing and terms.
The five ratings: output a model can produce 2/4 · needs a body in a room 0/4 · needs an accountable person 2/4 · needs to be trusted in the moment 1/4 · how much data exists 2/4.
Helping develop plans for new lines of business in collaboration with programmers and underwriters
staying human
The value here is that a specific person handles develop plans and stands behind it. That is earned, not computed.
importance 70 · 2433/01
Source: “Help develop plans for new lines of business in collaboration with programmers and underwriters.” (UK task statement)
How this row was scored
Exposure score: 31 out of 100 (24–38 allowing for uncertainty): low exposure, medium confidence.
Why it sits in this group: the value is that a specific person does it.
The rating behind it: Planning new business lines is collaborative work depending on colleagues’ commercial knowledge.
The five ratings: output a model can produce 2/4 · needs a body in a room 0/4 · needs an accountable person 1/4 · needs to be trusted in the moment 2/4 · how much data exists 2/4.
Evaluating and facilitating mergers and acquisitions processes
staying human
The value here is that a specific person handles mergers and stands behind it. That is earned, not computed.
importance 65 · 2433/01
Source: “Evaluate and facilitate mergers and acquisitions processes.” (UK task statement)
How this row was scored
Exposure score: 31 out of 100 (24–38 allowing for uncertainty): low exposure, medium confidence.
Why it sits in this group: the value is that a specific person does it.
The rating behind it: Deal work involves confidential judgement and a great deal of negotiation between people.
The five ratings: output a model can produce 2/4 · needs a body in a room 0/4 · needs an accountable person 1/4 · needs to be trusted in the moment 2/4 · how much data exists 2/4.
Recommending specific financial or operational strategies for client organisations to consider
staying human
The value here is that a specific person handles specific financial or operational strategies and stands behind it. That is earned, not computed.
importance 80 · 2433/03
Source: “Recommend specific financial or operational strategies for client organisations to consider.” (UK task statement)
How this row was scored
Exposure score: 31 out of 100 (24–38 allowing for uncertainty): low exposure, medium confidence.
Why it sits in this group: the value is that a specific person does it.
The rating behind it: Strategy recommendations depend on knowing the client organisation and earning its confidence.
The five ratings: output a model can produce 2/4 · needs a body in a room 0/4 · needs an accountable person 1/4 · needs to be trusted in the moment 2/4 · how much data exists 2/4.
Teaching courses in environmental economics
staying human
The value here is that a specific person handles courses and stands behind it. That is earned, not computed.
importance 60 · 2433/02
Source: “Teach courses in environmental economics.” (UK task statement)
How this row was scored
Exposure score: 30 out of 100 (23–37 allowing for uncertainty): low exposure, medium confidence.
Why it sits in this group: the same decision, made over and over; the value is that a specific person does it.
The rating behind it: Course material is easy to prepare, but teaching a class is live work with students.
The five ratings: output a model can produce 2/4 · needs a body in a room 1/4 · needs an accountable person 0/4 · needs to be trusted in the moment 2/4 · how much data exists 3/4.
Providing training to staff on applying mathematical models for data analysis
staying human
The value here is that a specific person handles staff and stands behind it. That is earned, not computed.
importance 80 · 2433/04
Source: “Provide training to staff on applying mathematical models for data analysis.” (UK task statement)
How this row was scored
Exposure score: 30 out of 100 (23–37 allowing for uncertainty): low exposure, medium confidence.
Why it sits in this group: the same decision, made over and over; the value is that a specific person does it.
The rating behind it: Materials can be written quickly, but training staff is live teaching with people in the room.
The five ratings: output a model can produce 2/4 · needs a body in a room 1/4 · needs an accountable person 0/4 · needs to be trusted in the moment 2/4 · how much data exists 3/4.
Guiding students' study projects to facilitate academic development
staying human
The value here is that a specific person handles students' study projects and stands behind it. That is earned, not computed.
importance 60 · 2433/02
Source: “Guide students' study projects to facilitate academic development.” (UK task statement)
How this row was scored
Exposure score: 28 out of 100 (21–35 allowing for uncertainty): low exposure, medium confidence.
Why it sits in this group: the value is that a specific person does it.
The rating behind it: Guiding a student’s project works through an ongoing relationship with that particular person.
The five ratings: output a model can produce 2/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 3/4 · how much data exists 2/4.
Collaborating with mathematicians to gather information for complex problem-solving
staying human
The value here is that a specific person handles mathematicians and stands behind it. That is earned, not computed.
importance 85 · 2433/03
Source: “Collaborate with mathematicians to gather information for complex problem-solving.” (UK task statement)
How this row was scored
Exposure score: 28 out of 100 (21–35 allowing for uncertainty): low exposure, medium confidence.
Why it sits in this group: the same decision, made over and over; the value is that a specific person does it.
The rating behind it: Working through hard problems with other mathematicians depends on live discussion between people.
The five ratings: output a model can produce 1/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 2/4 · how much data exists 3/4.
Reporting to and advising clients in consultancy roles
staying human
The value here is that a specific person handles and advising clients and stands behind it. That is earned, not computed.
importance 80 · 2433/01
Source: “Report to and advise clients in consultancy roles.” (UK task statement)
How this row was scored
Exposure score: 26 out of 100 (19–33 allowing for uncertainty): low exposure, medium confidence.
Why it sits in this group: the same decision, made over and over; the value is that a specific person does it.
The rating behind it: Reports can be drafted quickly, but consultancy advice rests on a relationship with the particular client.
The five ratings: output a model can produce 2/4 · needs a body in a room 1/4 · needs an accountable person 1/4 · needs to be trusted in the moment 2/4 · how much data exists 3/4.
Advising on selecting investment managers
staying human
The rules require a named, qualified person to answer for selecting investment managers, and that person cannot be a piece of software.
importance 75 · 2433/01
Source: “Advise on selecting investment managers.” (UK task statement)
How this row was scored
Exposure score: 26 out of 100 (19–33 allowing for uncertainty): low exposure, medium confidence.
Why it sits in this group: someone qualified has to answer for it; the value is that a specific person does it.
The rating behind it: Recommending investment managers is regulated advice built on trust with the client.
The five ratings: output a model can produce 2/4 · needs a body in a room 0/4 · needs an accountable person 2/4 · needs to be trusted in the moment 2/4 · how much data exists 2/4.
Advising fund managers on investment risks and payoffs for each recommendation
staying human
The rules require a named, qualified person to answer for fund managers, and that person cannot be a piece of software.
importance 85 · 2433/01
Source: “Advise fund managers on investment risks and payoffs for each recommendation.” (UK task statement)
How this row was scored
Exposure score: 25 out of 100 (18–32 allowing for uncertainty): low exposure, medium confidence.
Why it sits in this group: the same decision, made over and over; someone qualified has to answer for it; the value is that a specific person does it.
The rating behind it: Investment advice is a regulated activity in the UK, so an approved person must give it and answer for it.
The five ratings: output a model can produce 2/4 · needs a body in a room 0/4 · needs an accountable person 3/4 · needs to be trusted in the moment 2/4 · how much data exists 3/4.
Collaborating with internal and external stakeholders to facilitate reinsurance transactions and agreements
staying human
The value here is that a specific person handles internal and external stakeholders and stands behind it. That is earned, not computed.
importance 65 · 2433/01
Source: “Collaborate with internal and external stakeholders to facilitate reinsurance transactions and agreements.” (UK task statement)
How this row was scored
Exposure score: 24 out of 100 (17–31 allowing for uncertainty): low exposure, medium confidence.
Why it sits in this group: the value is that a specific person does it.
The rating behind it: Reinsurance deals are negotiated between people who need to trust each other.
The five ratings: output a model can produce 2/4 · needs a body in a room 0/4 · needs an accountable person 1/4 · needs to be trusted in the moment 3/4 · how much data exists 2/4.
Supervising economic research activities to analyse market trends
staying human
The value here is that a specific person handles economic research activities and stands behind it. That is earned, not computed.
importance 85 · 2433/02
Source: “Supervise economic research activities to analyse market trends.” (UK task statement)
How this row was scored
Exposure score: 23 out of 100 (16–30 allowing for uncertainty): low exposure, medium confidence.
Why it sits in this group: the value is that a specific person does it.
The rating behind it: Supervising researchers is about managing people, which software cannot take over.
The five ratings: output a model can produce 1/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 2/4 · how much data exists 2/4.
Assembling the right resources, including stakeholders and analysts, to deliver high-quality economic analysis for informed decision-making
staying human
The value here is that a specific person handles the right resources, including stakeholders and analysts and stands behind it. That is earned, not computed.
importance 80 · 2433/02
Source: “Assemble the right resources, including stakeholders and analysts, to deliver high-quality economic analysis for informed decision-making.” (UK task statement)
How this row was scored
Exposure score: 23 out of 100 (16–30 allowing for uncertainty): low exposure, medium confidence.
Why it sits in this group: the value is that a specific person does it.
The rating behind it: Pulling the right team together is about people and organisational know-how.
The five ratings: output a model can produce 1/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 2/4 · how much data exists 2/4.
Managing client expectations
staying human
The value here is that a specific person handles client expectations and stands behind it. That is earned, not computed.
importance 65 · 2433/02
Source: “Manage client expectations.” (UK task statement)
How this row was scored
Exposure score: 18 out of 100 (14–22 allowing for uncertainty): minimal exposure, high confidence.
Why it sits in this group: the value is that a specific person does it.
The rating behind it: Keeping a client comfortable with bad news or delay depends on a trusted personal relationship.
The five ratings: output a model can produce 1/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 3/4 · how much data exists 2/4.
Developing others through effective coaching
staying human
The value here is that a specific person handles this work and stands behind it. That is earned, not computed.
importance 60 · 2433/05
Source: “Develop others through effective coaching, mentoring, and training.” (UK task statement)
How this row was scored
Exposure score: 18 out of 100 (14–22 allowing for uncertainty): minimal exposure, high confidence.
Why it sits in this group: the value is that a specific person does it.
The rating behind it: Coaching and mentoring work through a trusting relationship with a particular person.
The five ratings: output a model can produce 1/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 3/4 · how much data exists 2/4.
Attending conferences and seminars for professional development
staying human
This work happens in the physical world: conferences, in a real place. Software cannot follow it there.
importance 60 · 2433/05
Source: “Attend conferences and seminars for professional development.” (UK task statement)
How this row was scored
Exposure score: 17 out of 100 (10–24 allowing for uncertainty): minimal exposure, medium confidence.
Why it sits in this group: the same decision, made over and over; work that happens in the physical world.
The rating behind it: Attending in person is the point of it, even where sessions are also available online.
The five ratings: output a model can produce 1/4 · needs a body in a room 2/4 · needs an accountable person 0/4 · needs to be trusted in the moment 1/4 · how much data exists 3/4.
Presenting recommendations at parliamentary hearings based on cost-benefit analyses
staying human
This work happens in the physical world: recommendations, in a real place. Software cannot follow it there.
importance 80 · 2433/02
Source: “Present recommendations at parliamentary hearings based on cost-benefit analyses.” (UK task statement)
How this row was scored
Exposure score: 16 out of 100 (9–23 allowing for uncertainty): minimal exposure, medium confidence.
Why it sits in this group: the same decision, made over and over; work that happens in the physical world; the value is that a specific person does it.
The rating behind it: Appearing before a parliamentary committee means a person answering live questions under scrutiny.
The five ratings: output a model can produce 2/4 · needs a body in a room 2/4 · needs an accountable person 0/4 · needs to be trusted in the moment 3/4 · how much data exists 3/4.
Negotiating changes to work scope or utilising networks to increase resources
staying human
The value here is that a specific person handles changes and stands behind it. That is earned, not computed.
importance 60 · 2433/02
Source: “Negotiate changes to work scope or utilise networks to increase resources.” (UK task statement)
How this row was scored
Exposure score: 14 out of 100 (7–21 allowing for uncertainty): minimal exposure, medium confidence.
Why it sits in this group: the value is that a specific person does it.
The rating behind it: Negotiating scope and calling in favours depends entirely on personal relationships.
The five ratings: output a model can produce 1/4 · needs a body in a room 0/4 · needs an accountable person 0/4 · needs to be trusted in the moment 3/4 · how much data exists 1/4.
Carrying out relationship management, liaising with clients, investment managers, finance directors and external stakeholders
staying human
The value here is that a specific person handles relationship management and stands behind it. That is earned, not computed.
importance 70 · 2433/01
Source: “Carry out relationship management, liaising with clients such as investment managers, finance directors and external stakeholders.” (UK task statement)
How this row was scored
Exposure score: 13 out of 100 (9–17 allowing for uncertainty): minimal exposure, high confidence.
Why it sits in this group: the value is that a specific person does it.
The rating behind it: Relationship management is built on trust with particular people and cannot be handed over to software.
The five ratings: output a model can produce 1/4 · needs a body in a room 1/4 · needs an accountable person 0/4 · needs to be trusted in the moment 3/4 · how much data exists 2/4.
Managing a research team, including technicians and supporting staff, to conduct studies
staying human
The value here is that a specific person handles a research team, including technicians and supporting staff and stands behind it. That is earned, not computed.
importance 75 · 2433/03
Source: “Manage a research team, including technicians and support staff, to conduct studies.” (UK task statement)
How this row was scored
Exposure score: 13 out of 100 (9–17 allowing for uncertainty): minimal exposure, high confidence.
Why it sits in this group: the value is that a specific person does it.
The rating behind it: Running a research team is about people, priorities and day-to-day judgement rather than producing a document.
The five ratings: output a model can produce 1/4 · needs a body in a room 1/4 · needs an accountable person 0/4 · needs to be trusted in the moment 3/4 · how much data exists 2/4.
Conducting clinical research focused on specific areas of medical interest
staying human
This work happens in the physical world: clinical research, in a real place. Software cannot follow it there.
importance 70 · 2433/05
Source: “Conduct clinical research focused on specific areas of medical interest.” (UK task statement)
How this row was scored
Exposure score: 10 out of 100 (3–17 allowing for uncertainty): minimal exposure, medium confidence.
Why it sits in this group: work that happens in the physical world; someone qualified has to answer for it.
The rating behind it: Clinical studies run under strict approvals with a named responsible investigator, and much of the work happens with patients.
The five ratings: output a model can produce 1/4 · needs a body in a room 2/4 · needs an accountable person 2/4 · needs to be trusted in the moment 1/4 · how much data exists 2/4.
Attending conferences, seminars and training sessions to stay informed about field developments
staying human
This work happens in the physical world: conferences, seminars and training sessions, in a real place. Software cannot follow it there.
importance 80 · 2433/03
Source: “Attend conferences, seminars, and training sessions to stay informed about field developments.” (UK task statement)
How this row was scored
Exposure score: 8 out of 100 (1–15 allowing for uncertainty): minimal exposure, medium confidence.
Why it sits in this group: the same decision, made over and over; work that happens in the physical world; the value is that a specific person does it.
The rating behind it: Attending conferences and training is about being there and meeting people in the field.
The five ratings: output a model can produce 0/4 · needs a body in a room 2/4 · needs an accountable person 0/4 · needs to be trusted in the moment 2/4 · how much data exists 3/4.
Acting as an expert witness during legal proceedings
staying human
This work happens in the physical world: an expert witness during legal proceedings, in a real place. Software cannot follow it there.
importance 60 · 2433/02
Source: “Act as an expert witness during legal proceedings.” (UK task statement)
How this row was scored
Exposure score: 7 out of 100 (0–14 allowing for uncertainty): minimal exposure, medium confidence.
Why it sits in this group: work that happens in the physical world; someone qualified has to answer for it; the value is that a specific person does it.
The rating behind it: Expert evidence must be given personally by a named individual who can be cross-examined and is answerable to the court.
The five ratings: output a model can produce 1/4 · needs a body in a room 2/4 · needs an accountable person 3/4 · needs to be trusted in the moment 2/4 · how much data exists 2/4.
Attending conferences to engage with the academic community
staying human
This work happens in the physical world: conferences, in a real place. Software cannot follow it there.
importance 75 · 2433/03
Source: “Attend conferences to engage with the academic community.” (UK task statement)
How this row was scored
Exposure score: 3 out of 100 (0–10 allowing for uncertainty): minimal exposure, medium confidence.
Why it sits in this group: work that happens in the physical world; the value is that a specific person does it.
The rating behind it: Conferences are about being in the room and meeting people, which software cannot do for you.
The five ratings: output a model can produce 1/4 · needs a body in a room 3/4 · needs an accountable person 0/4 · needs to be trusted in the moment 3/4 · how much data exists 1/4.
Performing laboratory tests to produce reliable data for scientific investigations
staying human
This work happens in the physical world: laboratory tests, in a real place. Software cannot follow it there.
importance 75 · 2433/04
Source: “Perform laboratory tests to produce reliable data for scientific investigations.” (UK task statement)
How this row was scored
Exposure score: 0 out of 100 (0–4 allowing for uncertainty): minimal exposure, high confidence.
Why it sits in this group: work that happens in the physical world.
The rating behind it: Laboratory testing is hands-on bench work with physical samples and instruments.
The five ratings: output a model can produce 0/4 · needs a body in a room 4/4 · needs an accountable person 0/4 · needs to be trusted in the moment 0/4 · how much data exists 2/4.

What this job pays, and how many people do it

Median pay: £53,300a year, before tax, the middle of the range, so half earn more and half earn less.ashe-t14, 2025 · ASHE 2025 provisional (reference April 2025)Provisional, because the ONS revises this figure in the autumn.
How we know this
Source: ashe-t14
Reference period: ASHE 2025 provisional (reference April 2025)
Rounding: Shown to the nearest £100. The exact published figure is in the downloadable dataset. We do not render pounds the survey cannot support.
People doing this job: 85,900in the UK, 2026.nomis-aps · Apr 2025-Mar 2026 (latest APS 12-month period)This headcount comes from a survey, not a census, so treat it as a good estimate rather than an exact count.

What is deliberately not here: a forecast of how many of these jobs exist in ten years. Where an official projection exists for a market we publish it with its vintage; where it does not, we leave the space empty rather than borrow the other country’s number.

Why this is shifting

The reason is boringly specific. Most of what is shifting here is reading one thing and writing another: data in, a record out. The rows above are exactly that shape: analysing data using mathematical and statistical methods to identify trends and analysing data to offer advanced insights and interpretations. What it cannot do is be trusted in person, which is what teams run on: someone specific doing it and standing behind it. Which is why this page talks about your tasks changing, not your job ending.

Your move

Over a pint: what I’d tell you if you were my friend

The exposed part of your job is the biggest part, and I am not going to dress that up: analysing data using mathematical and statistical methods to identify trends is work today's tools do quickly and cheaply, and that is not coming back.

So, given all that: 62% of this job's task weight sits in rows the software is already learning, 20% in rows that change shape rather than disappear, and 18% in rows it is nowhere near. That is the position, measured across 234 scored tasks. It is not a forecast about you.

What you have that the software does not is working collaboratively in teams and maintaining professional working relationships with colleagues and external stakeholders, plus the years of knowing when something looks wrong before you can say why. That is the raw material for everything below.

This week: one thing

Sit on the machine's side of the desk. Pick one real piece of data you would normally do yourself, let whatever software you already have take the first pass at it, and then go through what it produced and write down every single thing it got wrong. One evening this week. Do not fix anything yet. Just catch it.

What you end up holding: a written list of the machine’s mistakes, in your handwriting
How long it takes: an evening, or an hour if you pick one job rather than one client

If you can’t run software on your employer’s or your clients’ data: Do the same hunt on paper. Take one printed piece of data, work through it the way you always do, and mark every point where you made a call rather than followed a rule. Same evening, same list, nothing to log into and nobody to ask permission from. That list is the same artifact: it is the judgment written down.

Over the next 90 days

Change one sentence about what you do. Not on a CV. Out loud, to whoever pays you. From “my job is analysing data using mathematical and statistical methods to identify trends” to “I check what the software does and tell you what it means.” Your error list from this week is what makes that sentence true instead of a claim, so use it: show it once, to one person, inside the next ninety days. Same skills, priced as judgment rather than as typing.

Over the next 12 months

Walk toward the end of this job that answers for things, and get it recognised. Pick the one part of working collaboratively in teams and maintaining professional working relationships with colleagues and external stakeholders you are already best at, and spend the year making it formal: a qualification, a named responsibility, a specialism people ask for by name. Price it honestly: that is evenings, it is months rather than weeks, and the seats are competitive because everyone in your position is looking at the same door. Before you pay for anything, spend an hour with National Careers Service. It is free and government-funded, and it is listed below with the rest of the free routes.

The roads out of here, and why I am not sending you down them

I looked at the obvious moves out of this job, and here is what I found.

I checked the 12 nearest UK occupations to actuaries, economists and statisticians (nearest by the work that AI is not taking, not by job title), and none of them survived. The closest was data analysts: only about 2% of its durable work is work you already do, it is under the same pressure this job is and it pays 27.7% less. I am not going to pretend that is comfortable news: 62% of your own task list is already in the top exposure band. But the answer on this evidence is not a sideways jump into a job with the same problem. It is to walk toward the end of this one that answers for things. “ascertain premium rates required to ensure payment of future benefits” is the part that stays with a person, and spending a year getting formally recognised for it beats a standing start somewhere else.

How that was checked: this job was compared against all 412 UK occupations in this release on their official task statements, and the 12 nearest were examined one by one. A move that turns on an industry, an employer or a qualification rather than on the work itself will not show up in a check like that. And this release carries no licence register, so anything you are weighing needs that looked up separately.

3 moves I checked and rejected

These are the obvious-looking jumps. They are here with their reasons rather than quietly dropped, because the ones that fail are worth knowing about. It is one less thing to turn over at night.

Data analysts
Why it looked obvious: It came up as a near neighbour because one of your tasks is on their list in the same words: “analyse client needs using a range of Operational Research methods, adapting and developing them whilst understanding…”. Across the whole of both lists that adds up to about 2% of the work in that job the software is not taking.
Why I am not recommending it: Almost none of it is work you already do: about 2% of the durable side of that job. That is a different job, not a next step. I will not move you off one melting floe onto another: 74% of its own task list already scores in the top exposure band (71/100 in this release), so the same software is eating it. It is a pay cut, in those words: £38,572 against your £53,342, 27.7% less (ASHE Table 14.7a, 2025 provisional (both)). Retraining to earn less is a decision, not advice.
Look at that job’s page anyway →
Other researchers, unspecified discipline
Why it looked obvious: It came up as a near neighbour on the work AI is not taking: you already determine data collection methods for research projects or surveys, and their equivalent is to conduct fieldwork to collect data for research. Across both published task lists that is about 1% of the durable work in that job.
Why I am not recommending it: Almost none of it is work you already do: about 1% of the durable side of that job. That is a different job, not a next step. It is a pay cut, in those words: £44,580 against your £53,342, 16.4% less (ASHE Table 14.7a, 2025 provisional (both)). Retraining to earn less is a decision, not advice.
Look at that job’s page anyway →
Advertising and marketing associate professionals
Why it looked obvious: It came up as a near neighbour on the work AI is not taking: you already determine data collection methods for research projects or surveys, and their equivalent is to conduct fieldwork to collect data for research. Across both published task lists that is about 1% of the durable work in that job.
Why I am not recommending it: Almost none of it is work you already do: about 1% of the durable side of that job. That is a different job, not a next step. I will not move you off one melting floe onto another: 58% of its own task list already scores in the top exposure band (61/100 in this release), so the same software is eating it. It is a pay cut, in those words: £33,412 against your £53,342, 37.4% less (ASHE Table 14.7a, 2025 provisional (both)). Retraining to earn less is a decision, not advice.
Look at that job’s page anyway →

What I’d stop worrying about

A friend tells you what not to spend fear on. This is that list.

The headline number you read somewhere
The big “X% of jobs” figures are about the whole economy, not about you. The number that describes your job is on this page: 62% of its task weight, across 234 scored tasks. Every row behind it is printed above with the source. Check ours; ignore theirs.
“It’s too late for me to become something else”
You are not starting from zero, and the page shows why: working collaboratively in teams and maintaining professional working relationships with colleagues and external stakeholders is work the software cannot do and you already do it. The move above is a repricing of what you know, not a new career. Nobody who has just left college has that.
“I should learn to code”
Almost certainly not. The value in your job is moving toward checking, deciding and answering for the output, not toward writing the software. A weekend of Python will not change your position; the error list from this week will.
The “obvious” next job everyone suggests
I checked the obvious moves and most of them did not survive. The reasons are printed with the routes above, including the pay and the gate. A move that fails on the numbers is worth knowing about so you can stop turning it over at night.

You are reading the United Kingdom figures

The United States splits this work across more than one official group, of which Actuaries is the closest. The pay and employment figures are not directly comparable, and we do not average them together.

Switch to the United States page →partial match

The other groups this work is counted across:

In US official statistics this job is counted as Actuaries, Mathematicians, Operations Research Analysts, Statisticians, Economists and Survey Researchers. Pay is shown separately for each of those groups (medians cannot be averaged together), while the task list and the scores on this page are for this group only.

Your route through this

Where to go next, and what it costs

Free, and complete

The moves above cost nothing. These are the real services that go with them: public, government-funded, and free at the point of use. Nothing on this page is behind an email address or a payment.

In England:
National Careers Service
A free hour with a government-funded careers adviser is worth more than another evening of reading. In England that's the National Careers Service.
Free, government-funded
- Discover your skills and careers (the skills assessment)
In England:
National Careers Service - Find a course
Search what's actually running near you before you spend anything.
Free to search; individual courses vary
In England:
Free courses for jobs (Level 3 qualifications)
A free Level 3 qualification you already qualify for beats a paid course you don't need.
Free for eligible adults
In England:
Skills Bootcamps
Free, up to 16 weeks, and you get a job interview at the end. In England these are Skills Bootcamps - search what's running near you.
Free for eligible adults in England
- Skills Bootcamps - what they are and who can do them (DfE, Skills for Careers)
- Skills Bootcamps - JobHelp (DWP)
In Scotland:
My World of Work (Skills Development Scotland)
In Scotland it's My World of Work, from Skills Development Scotland.
Free, publicly funded
- Skills Development Scotland
In Wales:
Careers Wales
In Wales it's Careers Wales.
Free, Welsh Government-funded
In Northern Ireland:
Careers Service Northern Ireland
In Northern Ireland it's the Careers Service on nidirect.
Free, Department for the Economy-funded

No Space for this job, but one for what is happening to it

Nothing Collab365 runs is built for actuaries / economists / statisticians, and we are not going to point you at the nearest one and call it a fit.

There is one that is not about a job title at all. The AI Authority is about being the person who directs these tools at work rather than the person they get compared to. That is worth saying here, because 62% of the work on this page is already inside what they can do.

Try The AI Authority free →

7 days free, no card needed. Explore up to 2 Spaces before you choose a plan: you pick a plan later, not now.

The AI Authority is a general community about working with AI, not a course for actuaries / economists / statisticians. You do not need it to act on anything here: the moves above cost nothing and stand on their own. The data on this page is the same either way.

Noted, and thank you. We’ll email you if a Space for actuaries / economists / statisticians launches. Nothing else.

That did not look like an email address, so nothing was saved. Have another go below.

We could not save that. The fault is ours, not yours, and your address was not stored. Please try again later.

No Space for actuaries / economists / statisticians yet. Should there be one?

Collab365 Spaces is built by the same people as this site. We find the problems that AI and automation are creating inside one kind of work, then solve them as short courses, briefings and Blueprints. Each Space is the community too, so the research and the people doing your job are in the same place.

What a Space actually is, in full

Collab365 launches new communities where the need is real. If one for actuaries / economists / statisticians existed, with researched problems, courses and people in the same boat, would you want in?

Your email

We use your email address for one thing: to tell you if a Space for actuaries / economists / statisticians launches. We never sell it, never use it for unrelated marketing, and every email has a one-click unsubscribe. Our privacy policy has the full version.

This unlocks nothing. Every figure, every row and every step on this page is already yours, whether you fill this in or not.

No deadline on any of this. The page will still be here, and the data is refreshed on a published schedule rather than when someone wants a headline.

Questions people ask about this job

Will AI replace Actuaries, economists and statisticians?: Not as a job, but it is already doing parts of the work. Across the 234 official task statements scored for Actuaries, economists and statisticians (United Kingdom, SOC 2433), 62% of the importance-weighted core work is made of tasks today's AI could already do most of. The overall exposure score is 64 out of 100 (range 58–70, band: high). That is a statement about tasks, not about headcount: this measures what AI could do, not whether any employer adopts it, whether the law allows it, or whether doing the routine parts faster creates more demand for the human parts. Figures are from release 2026-q4.1.
Which tasks in “Actuaries, economists and statisticians” can AI already do?: The highest-scoring tasks in release 2026-q4.1 are: “Use programming languages and tools to manipulate and visualize data” (100/100, very high); “Write new functions or applications in programming languages to conduct analyses” (100/100, very high); “Implement data analysis algorithms” (100/100, very high). Each score comes from five published 0–4 ratings turned into a number by a published formula, and each carries the model's one-sentence reason on the page.
Which tasks in “Actuaries, economists and statisticians” stay human?: About 18% of this job's task weight sits in work that scores low for AI exposure. The lowest-scoring tasks in release 2026-q4.1 are: “Perform laboratory tests to produce reliable data for scientific investigations” (0/100, minimal); “Attend conferences to engage with the academic community” (3/100, minimal); “Act as an expert witness during legal proceedings” (7/100, minimal). Low scores usually mean the task needs a body in a room, a legally accountable human, or trust built in real time. Those are the three things the scoring rubric treats as gates rather than obstacles.
What should someone working in “Actuaries, economists and statisticians” do about AI?: Start from the ledger rather than the headline: 62% of this job's weighted core work is exposed, and roughly 18% is not. The practical move is to spend more of your week on the tasks that score low, the ones above, and to get fluent at directing AI through the tasks that score high, because those are the parts that change whether or not you are ready for them. This page does not predict your job, and nothing here is career advice tailored to you: the score describes the occupation, not the person.
How is the AI exposure score for Actuaries, economists and statisticians calculated?: Each official task statement for the occupation is rated on five published 0–4 dimensions (output replicability, physical embodiment, licensed accountability, real-time human trust, and data availability) by claude-opus-5 using scoring prompt task_scoring_v1.0. The model never writes the score; a published formula turns the five ratings into a 0–100 number, so every score can be recomputed by hand. The occupation figure is the importance-weighted mean across 234 scored tasks. The prompt, the rubric, the formula and the full dataset are published at https://futureproof.collab365.com/method and https://futureproof.collab365.com/data/2026-q4.1 under CC BY 4.0.

Where these numbers come from

About the data on this page

Provisional, because the ONS revises this figure in the autumn.
This headcount comes from a survey, not a census, so treat it as a good estimate rather than an exact count.
One row is marked low confidence, so treat it as a ballpark rather than a fine measurement.
The match between this job and its counterpart in the other country is partial, so the two pages count slightly different groups of people.
66 of this occupation's scored task statements carry a score that was measured under a different occupation's context, because the statement is shared between jobs and has only been scored once so far. Each one names the occupation it was measured under in the free routes below; none is presented as a measurement for this job.
Task clusters are not derived in this release, so the task-cluster field is empty and no Collab365 Space routing is attached to this occupation yet.

Task statements: gaisi-indexProcessing: catalogue-bridge → ssc-relatedness-weighting → task-scoring → score-aggregation
Task weights: gaisi-index (relatedness)
Scores: Rubric task_scoring_v1.0, model claude-opus-5, computed 2026-08-05.
Pay and employment: ashe-t14 (ASHE 2025 provisional (reference April 2025))nomis-aps (Apr 2025-Mar 2026 (latest APS 12-month period))

Figures on this page come from release 2026-q4.1, published 2026-08-05. Every release keeps its own permanent address, so a figure you cite in March is still there, unchanged, in November.

The plain-English wording on this page is assembled directly from the task statements and the published ratings, not written by hand for this occupation. That is why it is specific, and it is also why we say so.

The routes and free resources further up are today’s, not the release’s (last reviewed 2026-08-05). A route is an offer, not a historical fact, so it moves on its own clock.

How we score a job Download this release Look up another job

Using these figures?

Cite this

Everything on this site is published under CC BY 4.0. Quote it, chart it, sell something built on it. Just say where it came from, and cite the dated release rather than the site, so the figure you quote stays checkable.

Plain text

Collab365 (2026). Collab365 Futureproof: task-level AI exposure for US and UK occupations, release 2026-q4.1 (methodVersion 2.0.0, promptVersion task_scoring_v1.0). https://futureproof.collab365.com/data/2026-q4.1. Licensed CC BY 4.0. Built with O*NET data (USDOL/ETA, CC BY 4.0); ONS data (Open Government Licence v3.0); GAISI task framework (arXiv:2507.22748, MIT); BLS data (public domain).

BibTeX

@misc{collab365futureproof2026q41,
  title        = {Collab365 Futureproof: task-level AI exposure for US and UK occupations, release 2026-q4.1},
  author       = {{Collab365}},
  year         = {2026},
  url          = {https://futureproof.collab365.com/data/2026-q4.1},
  note         = {Release 2026-q4.1, methodVersion 2.0.0, promptVersion task_scoring_v1.0, CC BY 4.0}
}

Permalink

https://futureproof.collab365.com/data/2026-q4.1

Data as of release 2026-q4.1, published 2026-08-05. Releases never change after publication; when the figures move, a new dated release is published beside this one and this one stays exactly where it is.