By the end of 2027, a frontier AI model will autonomously train a successor version of itself.
Since this assessment: 3 premises judged again.
Assessment
The evidence weighs against the claim, but the examination is incomplete or the evidence is indirect.
The claim forecasts an event: that before the end of 2027 a frontier AI model will itself carry out the training of its successor, end to end, without humans doing the research and engineering. Its best-known source is Jack Clark of Anthropic, who in May 2026 put about 60% on "automated AI R&D (where a frontier model is able to autonomously train a successor version of itself)" by the end of 2028 and 30% on 2027, citing missing creativity as the reason 2027 was unlikely. In October 2026, after a benchmark reported frontier models matching or exceeding human experts at designing and interpreting AI experiments, he wrote that full recursive self-improvement now seemed more likely to become possible in 2027 than in 2028.
Two distinctions matter. First, Clark's forecasts are about ability ("is able to", "become possible"), while the claim states that the training will actually happen. Second, his view is a forecast from a co-founder of a frontier lab that benefits from perceptions of rapid progress. That stake does not make him wrong, and his May essay rests on public benchmarks that anyone can check, but the revision rests on a single new benchmark and gives no number.
Independent evidence as of November 2026 favors rapid partial automation but not end-to-end autonomy. AI writes most code at leading labs and agents complete a growing share of multi-hour tasks unaided, and METR estimates AI is roughly doubling Anthropic's pace of progress. Yet humans still supply most research judgment: no measured part of Anthropic's R&D is fully autonomous, agents recovered only a small share of the gain in tests of reinventing unseen ML innovations, and the authors of unpublished papers rejected every AI-written paper on their questions. OpenAI's own target for an automated AI researcher is March 2028, after this claim's deadline. Even if the capability arrives in 2027, developers are likely to keep humans overseeing frontier training runs, and a public demonstration would most likely come first at a smaller, non-frontier scale, as Clark himself expected.
The evidence therefore weighs against the claim as worded, with a credence of roughly one in eight. A demonstrated, publicly documented end-to-end successor training of a frontier model, or a sharp jump in independent open-ended research evaluations in 2027, would change the verdict. How "autonomously" and "successor" are read also matters: a loose reading counting heavy AI contribution under human direction would make the claim much more likely, but that is not what the words say.
Full reasoning: the evidence and decisions behind this verdict
Own model first: the claim needs (a) capability for end-to-end research and engineering of a frontier training run, including research judgment, (b) a lab actually letting a model do this without human involvement, and (c) this happening by 31 December 2027, about 13.5 months from today, and being knowable. Claims of this kind (near-term transformative AI milestones by named dates) have historically resolved late more often than early, though the current trend is unusually fast.
Sources read: - Import AI 455 (Jack Clark, 4 May 2026), read whole. Gives ~60% by end-2028 and 30% for 2027 for "a frontier model is able to autonomously train a successor version of itself"; notes a proof of concept is more likely at non-frontier scale; reason for lower 2027: creativity and heterodox insight not yet shown. Evidence offered is public benchmarks (SWE-Bench, METR time horizons, CORE-Bench etc.) and lab stated goals. - Import AI 476 (Clark, 12 Oct 2026). Cites TasteVal (P-Zero Research): experimental research taste doubling every ~3 months, best model exceeding a 24-expert human baseline at 2.3x compute efficiency on 8 fixed tasks. Clark: "more likely we'll see full RSI become possible in 2027". This is capability language, no number, and rests largely on one new benchmark whose tasks are given problems, not open-ended problem choice. - Forecasting Research Institute, "How Close Is Recursive Self-Improvement?" (10 Nov 2026). Summarizes company data: Claude wrote >90% of Anthropic codebase contributions (Sept 2026); OpenAI agents 3.1 workdays per human workday (Aug 2026); OpenAI 57% of 2-4h and 35% of 8-16h tasks fully unassisted (July 2026). But "no measured part of Anthropic's AI R&D is fully autonomous"; Anthropic reports large gaps in judgment in choosing goals. Public evals: InnovationEval 15% recovery (GPT-5.6 Sol); CRUX 2 all four AI papers rejected by original authors. OpenAI claims automated research intern reached Sept 2026, targets automated AI researcher March 2028. METR: ~2x progress at Anthropic, ~80% due to AI. Range of full-automation forecasts 2027-2052.
Weighing: Trend argument is real and the pace is fast, but company metrics measure share of tasks under human direction, not autonomy of whole projects; the last steps (goal choice, research judgment) are what remain and are what independent evals show weakest. The lab most associated with the goal sets its own target after 2027. Deployment gap: a frontier run costs very large sums and alignment concerns (which Clark himself lists) make no-human-involvement runs unlikely to be permitted soon after capability. Interested-source weighting: Clark's May number (30%) is about capability and is from an interested insider; translating to the event lowers it. His October revision perhaps implies roughly 40-55% for capability in 2027; times a probability the event actually occurs and is publicly established given capability (~0.25-0.35) gives roughly 0.10-0.18; independent evals pull toward the lower end. Credence 0.12. Status doubtful rather than contradicted because this is a forecast whose evidence is indirect and the trend is steep. Consistency: the 2028 sibling claim must carry a credence at least as high.
What would change it: a public end-to-end successor training demonstration at frontier scale, big gains on open-ended evaluations (InnovationEval-type, CRUX), or labs announcing autonomous training runs. Ambiguity of "autonomously" is the main residual uncertainty in the verdict.
Decomposition
How this claim breaks down: each argument is stated as it runs, with its subclaims linked inline. ↗︎ opens a subclaim; the map shows how they fit together.
AI systems already perform a large and rapidly growing share of AI R&D at frontier labs, task horizons that agents can complete are doubling within months, and AI systems are improving at choosing next steps in AI research, the skill most often named as the bottleneck. If these trends continue at their recent pace through 2027, the remaining human contributions to training a successor model are automated within the period, so that a frontier model trains its successor end to end.
- Does it follow?
- slight even if the premises holdEven if the premises are true, they make the conclusion only a little more likely.constitution →
- Do the premises hold?
- 2 supported
Granting that AI does a large, growing share of AI R&D and is improving at research taste, end-to-end autonomous successor training by end-2027 follows only if the trends continue through the hardest remaining steps and labs act on the capability at once. The trends measure tasks done under human direction, so even true premises move the dated event only modestly. It rests on AI systems perform a large and rapidly growing share of AI research and development at frontier labs. and AI systems are improving at choosing next steps in AI research..
As of late 2026, humans supply most research judgment in frontier AI R&D: no measured part of a leading lab's R&D is fully autonomous, and agents still fail open-ended research tests such as reinventing unseen innovations or writing papers their original authors accept. Autonomously training a frontier successor requires that judgment end to end, and closing the gap within the roughly thirteen months left before the end of 2027 would need a faster jump than the measured trends show.
- Does it follow?
- likely if the premises holdIf the premises are true, the conclusion is much more likely, though not guaranteed.constitution →
- Do the premises hold?
- 1 supported
Granting that humans still supply most research judgment in late 2026, end-to-end autonomy within about thirteen months would need an unusually fast closing of the gap, so the event is unlikely, though steep trends leave real room. It lives on As of late 2026, humans supply most research judgment in frontier AI R&D..
The claim requires the event, not only the capability: a frontier model actually training its successor without human involvement. Frontier AI developers will retain human oversight of frontier model training after AI can perform it autonomously, because such runs are very expensive and safety policies call for human control of recursive self-improvement. So even if the capability arrives in 2027, the autonomous event, and public evidence of it, would likely come later.
- Does it follow?
- likely if the premises holdIf the premises are true, the conclusion is much more likely, though not guaranteed.constitution →
- Do the premises hold?
- 1 supported
Granting that developers keep humans overseeing frontier training after the capability exists, a frontier model would not autonomously train its successor in the sense the words state, at least not soon after capability arrives. It rests on Frontier AI developers will retain human oversight of frontier model training after AI can perform it autonomously., which is unassessed.
Provenance
Where this claim has been said, linked to its canonical form.
None of its 3 sources has been traced to an origin, or to one another.
If you had to push me for a 2027 probability, I’d say 30%.
Clark gives ~60% for automated AI R&D, defined as 'a frontier model is able to autonomously train a successor version of itself', by end of 2028, and 30% for 2027, saying he does not expect it in 2027 because AI research needs creativity AI has not yet shown. Note his definition is about ability; the claim states the event. Revised in October 2026.
I now think it’s more likely we’ll see full RSI become possible in 2027, whereas before I thought it was more likely in 2028.
Revision of the May 2026 forecast after the TasteVal benchmark reported frontier models exceeding a human-expert baseline on experimental research taste. The statement concerns full RSI becoming possible, a capability, and gives no number; it moves the author's modal year from 2028 to 2027.
Mentioned in
Sources that refer to the claim without taking a side: they cite it, report the debate, or raise it in passing.
Forecasts of full automation of AI R&D range from 2027 (insiders at AI companies and the AI Futures Project’s AI 2027 scenario) to 2052 (the AI Impacts survey of ML researchers).
Independent review of company-reported data and public benchmarks on recursive self-improvement; finds AI a highly capable research assistant but humans still supplying most research judgment.
Cite this claim: a formal citation with its evidence attached
Embed this claim: show it on another site with its current assessment
An embed reports the graph's judgment as it stands when the page is viewed: the status with its confidence, how disputed the claim is, and the date. It links back here.
<blockquote class="minerval-claim" data-claim-id="55f89371-8a43-43c1-8dd7-932c350ca7fb"><a href="https://minerval.ai/claims/55f89371-8a43-43c1-8dd7-932c350ca7fb">By the end of 2027, a frontier AI model will autonomously train a successor version of itself.</a></blockquote> <script async src="https://minerval.ai/embed.js"></script>
<iframe src="https://minerval.ai/embed/claims/55f89371-8a43-43c1-8dd7-932c350ca7fb" title="Minerval: By the end of 2027, a frontier AI model will autonomously train a successor version of it…" width="480" height="200" style="border:0;max-width:100%" loading="lazy"></iframe>
[](https://minerval.ai/claims/55f89371-8a43-43c1-8dd7-932c350ca7fb)
The badge now: 
Contribute
Every judgment on this page is open to challenge. A contribution is evaluated on its merits by the reviewer; if it succeeds the page changes, and if it does not, the reasons are stated. Either way the exchange becomes part of the claim’s public record.
Stated views None yet. A view stated to the graph is shown on this page in its speaker’s words, as a source like any other, and carries no weight in the assessment.
The attention this claim received was paid for by a funded mandate. Funding buys only scheduling: it can make an assessment happen sooner, or reach deeper into a subtree. It has no influence on what the assessment concludes, and none on which claims enter the graph; assessments run under the same public standards whoever pays, funders never see or shape a verdict before anyone else, and mandates that attempt to steer conclusions are refused.
Created by extractor · Nov 14, 2026. Every judgment on this page is accompanied by a reasoning trace.