Skip to content

All five questions

Will AI actually boost productivity and growth?

The claim, as coded: “AI will substantially boost productivity and economic growth”

The short answer AI-written

Will AI substantially boost productivity and growth? Experiments show real task-level gains; the dispute is whether they scale. Recent macro estimates span roughly 0.25 to 1.3 percentage points of extra annual productivity growth, depending on assumed adoption and method. Optimists and skeptics include signatories of the letter, publishing in the same venues.

A synthesis of the cited research, not a prediction or a representative consensus.

Why they disagree AI-written

Published estimates of AI's yearly boost to productivity and growth run from under a tenth of a percentage point to several points, with some scenarios far higher.

Each row is one assumption and the published papers that make it (of 152 on this question). Dots: their estimates, hover for the paper. Bar: where their findings point.

What kind of evidence is the estimate built on?

Scaled-up task estimates and models16 papers · middle estimate 0.85
3 for, 10 mixed, 2 against
Measured in workplaces42 papers
No estimates
15 for, 19 mixed, 3 against
Expert judgment and surveys22 papers · middle estimate 0.66
7 for, 14 mixed

Would settle it: Studies that follow the same firms and sectors from task-level gains to measured output over several years.

How quickly will firms put AI to use?

Quickly4 papers
2 for, 1 mixed
Gradually or slowly7 papers
5 mixed, 2 against

Would settle it: Official firm surveys that track the share of work done with AI, and the returns firms report, year by year.

Does AI help workers or replace them?

Helps workers18 papers
No estimates
5 for, 12 mixed
Replaces tasks or whole roles10 papers
3 for, 7 mixed

Would settle it: Data on which tasks move from people to AI systems, and what happens to output when they do.

How much more capable will AI become?

About today's systems1 paper
1 for
Much more capable5 papers · middle estimate 20
2 for, 2 mixed

Would settle it: Measured progress of AI on long, real work tasks over the next few years.

The case made for the claim

Brynjolfsson has long argued AI is a general-purpose technology whose gains lag adoption, and co-leads a research agenda premised on transformative economic potential. Agrawal and Gans frame AI as a steep drop in the cost of prediction with broad productivity upside. Korinek models scenarios in which AI automating research itself produces explosive growth. Autor and Deming lean agree on the basis of AI's potential to extend expertise and its historically rapid adoption. Aghion and Bunel's two-method estimate puts AI's addition to productivity growth at 0.8-1.3 points a year by historical analogy, or a median 0.68 points of TFP growth under a task-based approach.

Still open

  • How much of the measured task-level gains survive workflow integration, hard task instances, and long horizons?
  • Will AI meaningfully accelerate R&D itself — the assumption doing the most work in explosive-growth scenarios?
  • How quickly do bottleneck sectors (the parts of the economy AI cannot speed up) come to dominate aggregate growth?
  • Why do macro estimates built from similar micro evidence differ by a factor of two or more, and which assumption (adoption pace, exposure, task difficulty) drives the gap?

Sources

  1. A Research Agenda for the Economics of Transformative AI
  2. The simple macroeconomics of AI
  3. Artificial Intelligence and Jobs: Evidence from Online Vacancies
  4. When Does Automating AI Research Produce Explosive Growth? Feedback Loops in Innovation Networks
  5. Artificial Intelligence and Economic Growth
  6. Generative AI at Work
  7. AI and Growth: Where Do We Stand?
  8. Miracle or Myth? Assessing the macroeconomic productivity gains from Artificial Intelligence
  9. Macroeconomic productivity gains from Artificial Intelligence in G7 economies
  10. Experimental evidence on the productivity effects of generative artificial intelligence

The numbers

Forecasts of AI’s effect on growth

Each point is one estimate of extra growth a year from AI, by the date it was published; bars show the author’s own range. 34 estimates from 32 sources, grouped as in Tom Cunningham’s Forecasts of AI & Economic Growth: economists and AI insiders. Measures (GDP, TFP, labour productivity), regions and horizons differ, and a few are converted from a level or a survey (the tooltip says how), so read the spread, not a single number. Log scale.

Economists: 0.3–0.85AI insiders: 4–200.010.030.10.313103010020172018201920202021202220232024202520262027pp a yearPwC (economist): +1 pp a year. Extra annual global GDP growth, implied by the 2030 level gain. “Global GDP will be up to 14% higher in 2030 as a result of the accelerating development and take-up of AI” Implied: 14% higher level in 2030, spread over the 13 years from 2017: 1.14^(1/13) - 1 = 1.0 pp a yearPwCBriggs & Kodnani (economist): +1.5 pp a year. Extra annual labour productivity growth. “~+1.5pp/yr US labor productivity”Briggs & KodnaniChui (economist): +0.35 pp a year (range 0.1 to 0.6). Extra annual productivity growth potential. “~+0.1-0.6pp/yr productivity growth potential”ChuiBaily et al. (economist): +2.7 pp a year. Extra annual growth in the Level+Growth scenario: the 18% level gain as a yearly rate plus 1 pp. “additionally includes a one percentage point boost in the rate of growth over the baseline rate” Implied: 18% higher level over ten years = 1.18^(1/10) - 1 = 1.7 pp a year, plus the one percentage point growth boostBaily et al.Cowen (economist): +0.38 pp a year (range 0.25 to 0.5). Extra annual US growth (best guess). “will boost the annual US growth rate by one-quarter to one-half of a percentage point” Implied: 'one-quarter to one-half' is written in words; value is the midpointCowenBergeaud (economist): +0.29 pp a year. Extra annual TFP growth. “equivalent to an additional 0.29 percentage points per year”BergeaudKorinek & Suh (economist): +18 pp a year. Steady-state output growth in the baseline AGI scenario (a total rate). “the economy remains in region 2 where labor and capital are perfectly substitutable, with steady-state growth of 18% per year”Korinek & SuhAldasoro et al. (economist): +2.7 pp a year. Extra annual GDP growth, implied by the 10-year level gain. “at which point GDP is almost 30% higher than it would have been without” Implied: GDP almost 30% higher after 10 years: 1.30^(1/10) - 1 = 2.7 pp a yearAschenbrenner (AI insider): +30 pp a year. Extra annual GDP growth from AI. “We could see economic growth rates of 30%/year and beyond, quite possibly multiple doublings a year.”AschenbrennerAghion & Bunel (economist): +1.05 pp a year (range 0.8 to 1.3). Extra annual aggregate productivity growth (past-revolutions approach). “the AI revolution should increase aggregate productivity growth by between 0.8 and 1.3pp per year over the next decade”Aghion & Bunel (economist): +0.68 pp a year. Extra annual TFP growth, median (task-based approach). “we obtain a median estimate of 0.68pp additional annual total factor productivity (TFP) growth”Tytell (economist): +0.25 pp a year (range 0.2 to 0.3). Extra annual productivity growth over the next decade. “a productivity boost of 0.2%– 0.3% is plausible over the next decade”Acemoğlu (economist): +0.066 pp a year (range 0.053 to 0.066). Extra annual TFP growth. “no more than a 0.66% increase in total factor productivity (TFP) over 10 years” Implied: 0.53-0.66% over 10 years divided by 10AcemoğluAmodei (Anthropic) (AI insider): +20 pp a year. Annual GDP growth in the developing world, a 'dream scenario' (a total rate). “would be 20% annual GDP growth rate in the developing world”Amodei (Anthropic)Filippucci et al. (economist): +0.65 pp a year (range 0.4 to 0.9). Extra annual labour productivity growth. “(0.4-0.9 pp. for labour productivity)”Filippucci et al.Filippucci et al. (economist): +0.42 pp a year (range 0.25 to 0.6). Extra annual TFP growth. “annual aggregate total-factor productivity growth due to AI range between 0.25-0.6 percentage points”Wiseman & McClements (AI insider): +6 pp a year (range 3 to 9). Extra annual economic growth from AI. “We expect AI will provide a 3%-9% increase to economic growth per year in the near future”Wiseman & McClementsCowen (economist): +0.5 pp a year. Extra annual growth. “AI will boost economic growth rates by half a percentage point a year” Implied: 'half a percentage point' is written in wordsMisch et al. (economist): +0.2 pp a year. Extra annual productivity growth. “at around 1 percent cumulatively over five years” Implied: 1 percent over five years divided by 5Clark (Anthropic) (AI insider): +4 pp a year (range 3 to 5). Extra annual growth, bear to bull case. “my bear case on all of this is 3 percent, and my bull case is something like 5 percent”Filippucci et al. (economist): +0.85 pp a year (range 0.4 to 1.3). Extra annual labour productivity growth. “annual aggregate labour productivity growth due to AI range between 0.4-1.3 percentage points in countries with high AI exposure”Filippucci et al. (economist): +0.5 pp a year (range 0.3 to 0.7). Extra annual TFP growth in the US. “AI could raise annual total factor productivity (TFP) growth by around 0.3-0.7 percentage points in the United States over the next decade”Wynne & Derr (economist): +0.3 pp a year. Extra annual productivity growth (their 'more reasonable scenario'). “AI boosts annual productivity growth by 0.3 percentage points for the next decade”Wynne & DerrArnon (economist): +0.15 pp a year. Extra annual growth to 2035. “AI will increase productivity and GDP by 1.5% by 2035” Implied: 1.5% by 2035 divided by about 10 yearsAnthropic (AI insider): +1.8 pp a year. Extra annual labour productivity growth from current models (extrapolation, not a prediction). “current-generation AI models could increase US labor productivity growth by 1.8% annually over the next decade”AnthropicIMF WEO (economist): +0.45 pp a year (range 0.1 to 0.8). Extra annual global growth in the medium term. “between 0.1 and 0.8 percentage points per year in the medium term”Jones & Tonetti (economist): +0.16 pp a year. Extra annual growth in GDP per person, implied (a model simulation, not the authors' forecast). “incomes are only 4% higher in 2050 than they would be if growth remained constant rather than accelerating” Implied: 4% higher income level in 2050, spread over about 25 years: 1.04^(1/25) - 1 = 0.16 pp a yearAmodei (Anthropic) (AI insider): +10 pp a year (range 5 to 15). Developed-world GDP growth he can see AI bringing (a total rate). “I can see a world where A.I. brings the developed world G.D.P. growth to something like 10, 15 percent. Five, 10, 15”CBO (economist): +0.1 pp a year. Extra annual TFP growth from generative AI. “annual TFP growth from 2026 to 2036 is 0.1 percentage point higher than it would have been without the additional contribution from generative AI diffusion”CBOKarger et al. (FRI survey) (economist): +0.5 pp a year. Extra annual GDP growth implied by the survey's median forecast over the baseline. “an annual GDP growth rate of 2.5%, which exceeds both the typical medium-run (2.0%) and long-run (1.7%) baseline forecasts” Implied: the median respondent's 2.5% a year minus the typical 2.0% medium-run baseline forecastLEAP experts (economist): +0.4 pp a year (range 0.4 to 1.7). Extra annual GDP growth implied by experts' forecasts over the slow-progress scenario (to rapid progress). “with median forecasts of 2.7%, 2.5%, and 2.5% annualized real GDP growth for 2025–2030, respectively—broadly in line with recent historical rates. Scenario assumptions make a substantial difference, however. Under Rapid Progress, the median expert expects 4.0% annualized growth to 2030 and 5.0% to 2050, while under Slow Progress those figures fall to 2.3% and 2.5%.” Implied: experts' unconditional 2.7% minus 2.3% under slow progress = 0.4; 4.0% under rapid progress minus 2.3% = 1.7Filippucci et al. (economist): +0.53 pp a year (range 0.1 to 0.95). Extra annual growth in per capita real income, across countries. “by 0.1–0.95 percentage points annually over the next decade in our central scenario”Brookings (economist): +3.63 pp a year. Extra annual GDP growth from AI. “A new BPEA paper projects that this buildout will total $10.3 trillion from 2025 to 2032, an average of 3.63% of U.S.”BrookingsAlrawashdeh et al. (economist): +0.83 pp a year. Extra annual GDP growth from AI. “The total effect is 0.829 at a significance level of 0.000, and the model accounts for 65.9% of the variance in economic growth.”

EconomistsAI insidersShaded: the middle half of each groupProductivity gain (including time saved)Open marker: AI labs’ own research

All 77 numbers from 52 papers, in 7 charts, each with its quote

Who says what

Documented positions only, coded with cited evidence by an AI agent: 26 of 400 mapped researchers so far. Everyone else is unknown: no reviewed evidence, which is neither agreement nor neutrality.

15 agree7 it's complicated4 disagree26 of 400 coded

Agree (15)

It's complicated (7)

Disagree (4)

Economists and computer scientists

Documented positions on this claim of researchers on the co-authorship map, by field (from their OpenAlex topics). Researchers without a documented position are not counted.

FieldAgreeIt’s complicatedDisagree
Economists (241 on the map)13 of 215 of 213 of 21
Computer scientists (51 on the map)Too few to say (3 coded)

Each position with the researcher’s reasoning and cited evidence

What to read

Essentials on productivity

Read

  • Experimental evidence on the productivity effects of generative artificial intelligenceShakked Noy and Whitney Zhang · Science · 2023 · Evidence for the claim

    A preregistered experiment with 453 professionals: ChatGPT cut writing time by 40% and raised quality by 18%.

  • Generative AI at WorkErik Brynjolfsson et al. · The Quarterly Journal of Economics · 2025 · Mixed evidence

    Gains measured in a real workplace: 15% more issues resolved per hour, large for novices, small for the most experienced.

  • The simple macroeconomics of AIDaron Acemoğlu · Economic Policy · 2024 · Evidence against the claim

    Why task-level gains may add up to little: at most 0.66% more total factor productivity over ten years.

Know

Follow

  • Erik Brynjolfsson agreesEconomist, Stanford University

    Co-author of the customer-support study and the productivity J-curve, which argues early gains are undercounted.

  • Daron Acemoglu leans disagreeEconomist, Massachusetts Institute of Technology

    Author of the task-based estimate that puts the ten-year productivity gain at no more than 0.66%.

  • Justified Posteriors (Andrey Fradkin)Newsletter · 7 posts in this guide

    Discusses new studies of AI and productivity as they appear.

All essentials on productivity, for 10 minutes to an afternoon

Where the papers point

49 agree156 it's complicated15 disagree220 of 313 coded

How each published paper’s own finding bears on the claim, read from its title and abstract by an AI model (AI-written). 220 of the 313 published papers address this question; the rest are not counted. This codes findings, not authors.

Findings that support the claim (49)

and 46 more in the research library

Findings that it depends (156)

and 153 more in the research library

Findings that go against the claim (15)

and 12 more in the research library

All research on this questionNews on this question