Anthropic co-founder Jack Clark moves recursive self-improvement forecast to 2027
Anthropic co-founder Jack Clark updated his artificial intelligence forecast, writing that full recursive self-improvement is now more likely to arrive in 2027 rather than 2028.
Jack Clark, co-founder of Anthropic, reported in his Import AI newsletter that he has shifted forward his expected timeline for artificial intelligence systems capable of recursive self-improvement (RSI). Clark wrote that he now views full RSI as more likely to become possible in 2027, revising his earlier projection targeting 2028.[1][3][4][6]
Clark grounded the revision in recent developments in machine learning capabilities and research creativity. Referencing his prior analysis in Import AI 455, Clark noted that while models had not demonstrated transformative creativity earlier in the year, recent milestones showed accelerated progress in creative-seeming mathematics breakthroughs and emerging proficiency in automated AI research benchmarks.[1][3][4]
The timeline adjustment has drawn attention across AI policy circles. Samuel Hammond of the Foundation for American Innovation argued that converging timelines from leading practitioners—including OpenAI's stated aim for an automated AI researcher by mid-2027—mean Washington's Super Intelligence Force cannot wait 120 days to take action as recursive AI advances.[5]
Key facts
- Anthropic co-founder Jack Clark stated that full recursive self-improvement (RSI) is more likely to become possible in 2027, revising an earlier forecast of 2028.
- Clark published the updated estimate in his Import AI newsletter, citing rapid improvements in AI creativity, math breakthroughs, and AI research abilities.
- Clark's previous May 2026 assessment placed roughly a 30% probability on RSI by 2027 and roughly 60% by 2028.
- Samuel Hammond cited Clark's revision and OpenAI's mid-2027 target for an automated researcher to argue that Washington's Super Intelligence Force must act immediately rather than take 120 days.
Sources · 5 sources
- CH
Chris@ChrisGPTPost on X ·
Jack Clark, Anthropic's co-founder, says he now believes full recursive self-improvement (RSI) is more likely to become possible in 2027 rather than 2028, citing rapid improvements in AI research capabilities and creativity. If you guys remember in his May 2026 analysis, he previously estimated a 30% probability by 2027 and roughly 60% by 2028.
Open source - MT
MTS@MTSlivePost on X ·
SITUATION DETECTED: Anthropic co-founder Jack Clark says he now believes it is likely that full recursive self-improvement will become possible in 2027.
Open source - CH
Chubby♨️@kimmonismusPost on X ·
Anthropics' Jack Clarke has adjusted his timeline for RSI and brought its arrival forward; he considers it highly likely that "full RSI will become possible in 2027, whereas before I thought it was more likely in 2028." Again, 2027 will be the year everything will drastically change forever. 2026 will be the last "normal", although noting feels normal any longer.
Open source - RW
Rob Wiblin@robertwiblinPost on X ·
Jack Clark of Anthropic now thinks it's more likely we see full RSI become possible in 2027 than in 2028. https://t.co/ttga3JItKY
Open source - MT
MTS@MTSlivePost on X ·
FAI's @hamandcheese says Washington's Super Intelligence Force can't afford to wait 120 days when recursive AI self-improvement could arrive as early as 2027: "I hope that they are not just using the whole 120 days, ’cause also in the news today was Jack Clark revising his timeline to fully automated AI R&D, recursive self-improvement, from like mid ‘28 to like Q1 ‘27." "I am a mere economist. I try to be a good Bayesian and defer to people who are closer to the bare metal. And there’s a striking convergence in people’s timelines. OpenAI talks about having their automated AI researcher by mid next year." "Do we go foom? Is it a 2 to 3X speed up? Even if it fizzles, it’s still a step change in the pace of acceleration. And so 120 days is like a lifetime. We need to start acting now, and this SI force needs to be hitting the ground running." "They get the companies together. They all sign this big commitment to monitoring, control, embedded evaluators. And then a day or two later, the FTC launches a probe into Anthropic, OpenAI, suggesting that they are really going to try to enforce this through the backdoor of FTC authority." @JoinFAI
Open source - TE
Techmeme@TechmemePost on X ·
Anthropic co-founder Jack Clark says "I now think it's more likely we'll see full RSI become possible in 2027", earlier than his previous prediction of 2028 (@jackclarksf / Import AI) (Visit Techmeme dot com for the link and full context!)
Open source