AI and jobs: Who expects how much human work to disappear?

International Labour Organization · Mostly task change through 2030

The ILO finds broad GenAI exposure but says human input makes transformation more likely than redundancy.

most jobs will be transformed rather than made redundant

International Labour Organization · Severe displacement within five years

The study points away from wholesale redundancy but gives no unemployment threshold or comparable aggregate forecast.

Global employers · Future of Jobs Survey 2024 · Severe displacement within five years

AI-related technologies are projected to create 11 million jobs and displace 9 million globally by 2030.

Global employers · Future of Jobs Survey 2024 · New work offsets losses

The forecast gives 11 million jobs created versus 9 million displaced, despite substantial gross churn.

expected to create 11 million jobs, while simultaneously displacing 9 million others

Erik Brynjolfsson · Mostly task change through 2030

Brynjolfsson argues for analysing tasks, with displacement concentrated where AI substitutes rather than complements workers.

Almost no job is fully automatable

Erik Brynjolfsson · Severe displacement within five years

He finds no economy-wide displacement so far, alongside a 19% gap for young workers in highly exposed occupations.

Erik Brynjolfsson · New work offsets losses

Firms often move workers to higher-value tasks, but the cited source does not quantify decade-long net job creation.

Dario Amodei · Mostly task change through 2030

Amodei says rapidly improving AI can close task gaps instead of remaining a complement.

Dario Amodei · Severe displacement within five years

He warned that about half of entry-level white-collar jobs could disappear, with unemployment potentially rising sharply.

AI could wipe out half of all entry-level white-collar jobs

Dario Amodei · Most paid work optional within 20 years

AI may technically do most jobs within years; deployment and physical-world automation can lag.

Dario Amodei · New work offsets losses

He argues advanced AI can also perform jobs that would normally emerge after automation.

a general labor substitute for humans

Sam Altman · Mostly task change through 2030

Altman says he overestimated near-term white-collar losses but gives no 2030 task-versus-job forecast.

Sam Altman · Severe displacement within five years

Altman explicitly moved away from his earlier fear of sweeping near-term white-collar elimination.

Jensen Huang · Mostly task change through 2030

Huang treats AI mainly as a productivity tool and points to occupations where demand expands.

Jensen Huang · Severe displacement within five years

Huang calls such mass-job-loss forecasts implausible and expects productivity to expand useful activity.

Jensen Huang · Most paid work optional within 20 years

On Musk’s 2025 panel, Huang forecast busier, more productive workers near term but made no matching long-run claim.

Jensen Huang · New work offsets losses

Lower costs and higher productivity, he argues, expand activity rather than simply cut headcount.

increasing the amount of jobs that the world needs

Elon Musk · Most paid work optional within 20 years

Musk expects AI and large-scale robotics to make paid work economically unnecessary for most people.

My prediction is that work will be optional.

Bill Gates · Mostly task change through 2030

Gates expects AI to make expert services abundant without specifying whether affected occupations mainly change or disappear.

Bill Gates · Most paid work optional within 20 years

Gates predicts AI will make most utilitarian human work unnecessary while leaving some activities to human choice.

Not for most things.

Trades Union Congress · Mostly task change through 2030

The TUC expects jobs to evolve as tasks change, while some groups still face displacement and deskilling.

Trades Union Congress · Severe displacement within five years

The TUC accepts concentrated displacement but rejects a near-term economy-wide collapse as the balanced reading.

a ‘rapid collapse’ in employment is unlikely

McKinsey Global Institute · Mostly task change through 2030

Automation cuts work hours, but demand growth can keep many workers in changed versions of their occupations.

McKinsey Global Institute · Severe displacement within five years

MGI expects more US jobs by 2035 but warns that timing mismatches can still create displacement.

McKinsey Global Institute · New work offsets losses

About 36 million job-equivalents of reduced demand are offset by more than 40 million jobs of growth.

The next decade’s challenge is mobility, not scarcity.

Kristalina Georgieva · Mostly task change through 2030

Georgieva sees new tasks and occupations emerging alongside automation, without saying which effect dominates.

Kristalina Georgieva · Severe displacement within five years

The IMF analysis finds broad exposure and some employment weakness, but no double-digit unemployment projection.

Kristalina Georgieva · New work offsets losses

New roles appear alongside automation; the analysis does not say they will compensate most losses.