How long until AI automates all cognitive labor?futuresearch☰SolutionsPricingResearchDocsEvalsBlogCompanyTry it for free← Back to BlogHow long until AI automates all cognitive labor?April 12, 2026By Dan SchwarzChanges in top researchers' timelines for when AI will automate all cognitive labor, 2023-2026Median AGI forecasts→→→~2023Early 2025Late 2025Apr 20262026202820302032203420362038204020422044204620262028203020322034203620382040204220442046D. KokotajloD. AmodeiN. JurkovicR. GreenblattD. HassabisFutureSearchT. LarsenE. LiflandY. BengioB. ToddA. CotraP. WildefordMetaculusT. BesirogluE. Erdil

Predictions for when most cognitive labor will be fully automated. Icons are medians, with approximate confidence intervals.
Follow up post to: AI 2027, One Year Later
Many orgs track AGI timelines, like AI Futures, Metaculus, and Epochs. Recently, though, I noticed that many great researchers have now published two or more precise forecasts, all using similar definitions of AGI, and all providing confidence intervals. So I was able to visualize how their forecasts changed over time.
(Most conventionally famous people in AI don't make specific predictions, and when they do, they don't generally update them, so we can't see their views changing over time. I actually prefer to learn from the people shown here, though, as I think they have the best track records.)
The overlapping AGI definition I use here is "Most purely cognitive labor is automatable at better quality, speed, and cost than humans". For some of these researchers, saying they use this definitions is a bit of a stretch, but I included everyone who I judged as close enough to be informative.
So now I could ask, are the best AI forecasters updating the same way that I wrote about last week, how Daniel Kokotajlo and Eli Lifland pushed their AGI timelines out during 2025, but then pulled them back in early 2026 given the rapid progress from Anthropic?
If you look at the graphic, you'll see that from 2023 to 2025, most people brought their AGI timelines in to be sooner, though with some exceptions like Tamay Besilogru. From 2025 to 2026, joining Daniel and Eli in pushing their timelines out are the Metaculus community, Dario Amodei, and elite forecast Peter Wildeford. In fact, across 2025, only Benjamin Todd brought in his timelines to say AGI would happen sooner.
(The graphic also rounds the date of the forecasts made to 4 points, for simplicity of seeing the patterns, so apologies to forecasters who are shown as making, for example, a "late 2025" update that actually did made in January 2026.)
And every single person who updated their timelines from January 2026 to April 2026 has moved their timeline to say AGI is coming sooner, myself included.
So I think the data supports the impression I got from Daniel, Eli, and the AI Futures team. One way I could characterize it is: in the ChatGPT era, people updated towards AI coming sooner. Then in the xAI, Meta, and Gemini era, people updated towards it coming later. Then in the Anthropic era, people updated towards AI coming sooner. Take from that what you will.
Good Bayesians shouldn't be able to predict which direction they will update. When I have intuitions about how I or others may update soon, it's useful to interpet that as evidence that I should update now. I kind of feel like I know what direction I'm going to update next, so...View All PostsGeneral inquiry? You can reach us at hello@futuresearch.ai.CompanyTeam & CareersPressPrivacy PolicyTerms of ServiceDevelopersSDK DocsAPI ReferenceCase StudiesGitHubSupportIntegrationsClaude CodeCursorChatGPT CodexClaude.aiFollow UsX (Twitter)@dschwarz26LinkedIn

Researchers tracking Artificial General Intelligence timelines have published various forecasts, and a recent analysis of these predictions reveals shifts in timelines correlating with the evolution of specific AI eras. The basis for these forecasts often relies on the definition that most purely cognitive labor is automatable at a better quality, speed, and cost than humans. Some researchers included in this cohort include D. Kokotajlo, D. Amodei, N. Jurkovic, R. Greenblatt, D. Hassabis, T. Larsen, E. Lifland, Y. Bengio, B. Todd, A. Cotra, P. Wildeford, T. Besiroglu, and E. Erdil.

An examination of the data shows how these predictions have evolved over time, particularly when segmenting the forecasts based on the context of the dominant AI development era. From 2023 to 2025, most participants tended to bring their estimated Artificial General Intelligence timelines forward, although exceptions existed. Subsequently, between 2025 and 2026, the Metaculus community, alongside researchers such as Dario Amodei, and Peter Wildeford, pushed their timelines further out. This collective movement indicates that the perception of when AGI will occur is heavily influenced by the rapidly changing landscape of AI development.

The pattern observed suggests a shift in forecasting direction based on the prevailing technological context: in the ChatGPT era, the collective tendency was to push timelines sooner; in the xAI, Meta, and Gemini era, the forecasts shifted toward a later timeline; and in the Anthropic era, the direction again favored timelines coming sooner. This dynamic illustrates that the predictive stance of researchers is responsive to recent AI progress.

The analysis further suggests that individuals should consider their own intuition about future updates as evidence to adjust their timelines. This perspective implies that understanding the direction in which future updates are likely to occur is beneficial for deciding when to revise one's forecasts.