Hiring nowAI economics, labour and social research

AI Forecaster

Also called Capabilities Forecaster

An AI Forecaster estimates when AI capabilities will arrive and what risks follow. Their judgement helps a company or policy body plan, set priorities and make decisions under real uncertainty.

What they do

The forecaster turns vague questions about AI progress into specific, checkable predictions. They break a question into parts, gather evidence from benchmarks, evaluations, compute trends and economic data, and build a view with stated uncertainty. They write down their reasoning, record probabilities and update as new evidence lands. They run or join forecasting exercises, compare their calls with outcomes, and work out where they were wrong. They also brief leaders and policy teams in plain language, explaining what a forecast means for a decision and what would change it.

What a strong hire delivers

In the first months, a strong hire sets up a clear forecasting process the team can reuse. They deliver a set of well-defined questions with written probabilities and reasoning, and a record that tracks accuracy over time. They produce briefings that leaders can act on, with the assumptions visible. They also help the team see which uncertainties matter most to strategy, and which are noise.

How you can tell they are good

Ask them to show past forecasts, including ones they got wrong. A strong candidate gives probabilities, not hedged prose, and can explain how they updated. Give them a fuzzy question about AI capabilities and watch them sharpen it, split it up and name the evidence they would seek. Worry if they sound certain, or cannot say what would change their mind.

Why it is hard to fill

Almost nobody carries this title, so you cannot search for it. The skill sits between quantitative reasoning, AI knowledge and social science, and few people have all three. Good forecasters often come from research or evaluation work and have a track record that is hard to see from a CV. Demand is high, because labs, public bodies and funds all want the same small group, and at least one major regulator is hiring forecasters openly.

Where to find them

Look in economics, sociology, labour research and evaluation science, where people already reason about causes and evidence. Strong candidates also come from forecasting tournaments, think tanks, policy units and AI safety research groups. Typical employers today are frontier labs, AI safety and security institutes, and funds. Titles include research analyst, economist, evaluations researcher, strategist and capabilities forecaster.

Indicative pay

Not enough published evidence yet to give a fair range for this role, which is common for work this new. We benchmark it for you as part of a search.

Interview questions for an AI Forecaster

  1. Walk me through a forecast you made about technology or work. How did you set it up and what happened?

    A strong answer names a specific question, gives probabilities, explains the evidence used and admits how it turned out. Worry about vague predictions, no record of results, or a story that only shows the candidate being right.

  2. How would you turn the question 'when will AI automate most research tasks?' into something you can forecast?

    They should define terms, set resolution criteria, split the question into measurable parts and name evidence sources. Worry if they answer with a single date or skip the definitions.

4 more questions, with what a strong answer covers, are in the Role Kit.

Job description you can post

We are hiring an AI Forecaster to help us anticipate how AI capabilities will develop and what that means for work, markets and society. You will turn uncertain questions into clear, tracked forecasts that shape our strategy and the advice we give others.

What you will do

  • Define forecasting questions with clear resolution criteria
  • Build and record probabilistic forecasts with written reasoning
  • Gather and weigh evidence from evaluations, economic data and research
  • Track accuracy and review where forecasts went wrong
  • Brief leaders and policy colleagues in plain language
  • Design simple studies to test key assumptions about AI and work

What you will bring

  • A record of forecasts with stated probabilities and results
  • Strong quantitative reasoning and comfort with uncertainty
  • Working knowledge of AI capabilities and how they are evaluated
  • Experience with causal inference or experimental design
  • Ability to write clearly for non-specialist decision makers
  • Understanding of public policy or economics

Nice to have

  • Experience in forecasting tournaments or similar exercises
  • Background in labour research or evaluation science
  • Experience working with workers or communities affected by technology

Free to use and adapt. If it saves you time, a link back to this page is appreciated.

Skills that matter

  • Public policy and economicsWorkingAnalyse institutional choices and distributional effects
  • Causal inferenceWorkingSeparate causal claims from predictive associations
  • Experimental designWorkingChoose comparisons that identify credible effects
  • Worker and community participationWorkingInclude affected workers and communities in decisions

Other roles in aI economics, labour and social research

Hiring nowAI Labour EconomistAn AI Labour Economist measures how AI adoption changes work and employment, and says how sure we can be. Companies hire them to ground product…Hiring nowAI Social Impact ResearcherAn AI Social Impact Researcher studies how AI products and policies affect communities, workers and institutions. They give the company evidence it…Hiring nowAI Workforce Planning AnalystAn AI Workforce Planning Analyst turns shifting task needs into clear workforce scenarios. It helps a company decide which skills to build, buy or…EmergingAI Market Structure ResearcherAn AI Market Structure Researcher studies who holds power in AI markets and why. They turn questions about concentration, access and competition…Hiring nowAI EconomistAn AI Economist measures what AI does to work, markets and prices. They give the company evidence it can use to set pricing, shape policy…

Related roles in other families

Hiring nowCausal Inference ScientistCausal inference, decisions and experimentationA Causal Inference Scientist estimates what actually changes when the company acts, under assumptions that are stated and tested. That keeps…Hiring nowExperimentation Data ScientistCausal inference, decisions and experimentationAn Experimentation Data Scientist designs and analyses product and operational experiments so the company can tell what actually works. They turn…Hiring nowDecision ScientistCausal inference, decisions and experimentationA Decision Scientist turns uncertain evidence into clear, constrained choices. They help an AI company decide what to ship, test, fund or stop, and…Hiring nowAlgorithmic Game Theory ResearcherCausal inference, decisions and experimentationAn Algorithmic Game Theory Researcher studies how agents respond to incentives inside computational systems. They design mechanisms and strategic…

How we wrote this

Role definitions are our own editorial work, checked against occupational classifications, standards, research and employer pages. Titles vary by company, and this is a description of the work, not a job advert.

← AI economics, labour and social researchAll new job families

Start a search

Hiring an AI Forecaster?

Tell us the work. We find the people already doing it, whatever their job title says, and you meet them already interviewed for your role.