📊 Full opportunity report: When AI Builds Itself: Inside Anthropic’s Evidence on Recursive Self-Improvement on ThorstenMeyerAI.com — validation score, market gap, and execution plan.
TL;DR
Anthropic’s new report provides data indicating that AI models are progressively automating tasks involved in AI development. While current evidence shows significant progress, full recursive self-improvement remains unconfirmed and is not yet inevitable.
Anthropic’s new report states that AI models are already increasingly automating parts of AI research and development, with some tasks now largely handled by AI systems themselves. This development raises questions about the potential for AI to eventually improve itself in a loop driven by compute power rather than human effort, although experts emphasize that full recursive self-improvement is not yet here.
The report from The Anthropic Institute presents concrete data showing that AI models like Claude now perform a growing share of coding and experimental tasks traditionally done by humans. For example, as of May 2026, over 80% of code merged into Anthropic’s codebase was authored by Claude, up from single digits in early 2025. Public benchmarks such as METR also demonstrate that AI’s ability to handle complex tasks has doubled roughly every four months, with models capable of completing tasks that previously took hours now handling 12-hour workloads.
Inside the labs, Anthropic’s analysis indicates that AI systems are improving in their ability to execute experiments and generate code, but the critical step—AI choosing which problems to pursue or which experiments to run—is still predominantly human-controlled. The authors highlight that AI can match or outperform skilled humans at executing specific, well-defined experiments, but remains weak at setting research goals or designing new approaches without human guidance.
When AI builds itself
Anthropic is delegating a growing share of AI development to AI. Taken far enough, that points to a system that designs its own successor — recursive self-improvement. Not here yet, not inevitable. But the case isn’t speculation: it’s data on what AI is doing to AI development right now.
The curve that hasn’t bent
METR tracks the length of tasks AI can reliably complete on its own. That horizon is doubling roughly every four months — up from every seven. Anyone can check this in public data.
Task horizon — how long a job AI can handle solo
Each model handles dramatically longer tasks than the one a year before. The line keeps going up.
AI VoiceWriter – Smart Dictation & AI Writing Assistant for Windows & Mac | USB Dongle & Mobile App for Voice Input, Proofreading, Rewriting & Multilingual Support
🎙️ Hands-Free Voice Typing for Windows & Mac – Powered by iOS & Android dictation technology, AI VoiceWriter…
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Two kinds of work, one persistent gap
Building a frontier model splits into engineering and research. Across both, the pattern is the same — and so is the one thing AI still can’t do well.
Code, infrastructure, training
Claude can take an underspecified problem and find a method. Humans supply the goal; they no longer need to supply the method.
Which experiments, what they mean
Claude can match or outperform skilled humans at executing a well-specified experiment. But choosing which experiment still needs a human.
The same ladder Anthropic employees climb with experience
The No-BS Guide to AI for Trading & Market Research: How to Use ChatGPT, Claude & AI Tools for Market Analysis, Stock Research & Data-Driven Trading … … Required (The No-BS AI Playbooks Book 3)
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Watch the human share shrink, rung by rung
Walk up the four stages of AI development. At each, the human/AI split shifts — and the real internal numbers show exactly where AI has reached parity, gone superhuman, or still trails. Tap a rung.
The human role across the development loop
The doing now costs almost nothing in human time. What’s left is the deciding.
machine learning experiment platform
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Agents ran an open research project end to end
April 2026: the first demonstration of Claude running an open-ended research project from hypotheses to findings — on a real AI-safety problem.
Can a weaker model reliably supervise a stronger one?
Agents were left to solve it: proposing hypotheses, testing them, sharing findings across parallel agents, iterating. Measured against the gap between a “floor” (weak supervisor alone) and “ceiling” (strong model trained on correct answers).
(humans: ~23% in a week)
· ~$18,000 compute
the agents themselves
AI development environment
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Picking a better next step than the human
Real research sessions where a human took a wrong turn. Models saw only the work before the detour and proposed a next step; a judge that knew the outcome scored them. The day-to-day of research is this chain of next-step calls.
“Can the model pick a better next step than the human?”
Share of moments where the model’s next step was judged better. The amber line is the practical ceiling (an ideal answer that could see the whole session).
It depends on whether the trend continues — and what we do
The piece refuses a single prediction. It lays out three scenarios, and is clear about which it finds most likely.
The exponentials turn out to be S-curves
Maybe taste can’t be scaled into existence; maybe the constraint is the supply chain — chips, grid, interconnect — not intelligence. Even so, the world still changes: Glasswing’s Mythos found 10,000+ critical vulnerabilities in weeks, and a 100-person firm does the work of 1,000.
included for completeness · they doubt itDevelopment automates; humans still steer
100-person companies doing the work of tens of thousands — revolutionary, but turnable to harm (population-scale surveillance, tailored manipulation). Bound by Amdahl’s law: speeding one part shifts the bottleneck — which is exactly why human code review became Anthropic’s new chokepoint.
★ they think we’re likely heading hereAI designs and refines its own successors
Progress paced only by compute. Humans move to oversight of an expanding “virtual lab.” The future they understand least — especially whether alignment holds, or whether rare misalignments compound as models build successors, until control slips.
the one they’re most uncertain aboutBuild the option to slow down — verifiably
The piece ends on policy, not product. A unilateral pause just changes who leads; what’s missing is the ability to verify others have actually slowed.
Why a credible pause is hard — and worth building toward
A slowdown that only lets the least cautious catch up leaves everyone less safe. So the goal is the option: systems that let frontier labs verify others have genuinely stopped. Anthropic says if such systems existed and peers paused verifiably, it expects it would too.
Detection beats verification — and even that’s tough
Training runs are easier to conceal than missile silos, inputs are general-purpose, and whoever continues while others pause inherits the lead.
We’ve done it before — slowly
Regimes like the INF Treaty built verification and trust over decades. The authors’ blunt line: “We don’t have that long.”
Reading it in proportion
- This is one lab’s account of its own internal data — much previously unreported, not independently audited.
- The soft spots are stated in the original: lines-of-code overstates productivity; the self-reported 4× is probably high; the headline research result didn’t transfer to production scale; the next-step test used cherry-picked moments.
- “More autonomous” is not “fully autonomous” — every standout result still had a human framing the problem and defining success.
- That the authors surface these caveats themselves — against their own incentive — is part of what makes the document serious.
Implications of Accelerating AI Self-Development
This evidence suggests that AI is rapidly advancing in automating parts of its own research process, which could lead to faster development cycles and possibly trigger recursive self-improvement if the remaining human decision points are automated. Such a shift could dramatically accelerate AI progress, raising questions about control, safety, and the pace of technological change.
However, experts caution that the leap from improved automation to full self-improvement is not yet confirmed. The key bottleneck—AI’s ability to autonomously decide which problems to solve—remains unresolved, and current data does not indicate that this is imminent. Still, the trend points to a future where AI could play a much larger role in shaping its own evolution, potentially at speeds that outpace human oversight.
Current Evidence of AI Progress in Development Tasks
The report builds on public benchmarks like METR, SWE-bench, and CORE-Bench, which track AI capabilities in coding, bug fixing, and research reproduction. These benchmarks show a consistent pattern: AI capabilities have been doubling every few months, with models now able to handle tasks that once required days or weeks of human effort. This acceleration aligns with internal data from Anthropic, which reveals that AI models are increasingly contributing to codebases and experimental workflows.
Prior to this, AI progress was often viewed as steady but incremental. The recent data, however, suggests a sharper upward trend, driven by larger models and improved training techniques. The report emphasizes that while public benchmarks measure task performance, they do not directly capture the internal pace of AI-driven research, which is where the most significant changes are occurring.
“Our data shows that AI models are now performing a substantial portion of the coding and experimental work involved in AI research, and this trend is accelerating.”
— Thorsten Meyer, lead author of the report
Unconfirmed Potential for Fully Autonomous Self-Improvement
It remains unclear whether AI will soon be capable of autonomously setting research goals and designing its own successors without human input. The report emphasizes that this step is not yet achieved, and whether it will occur depends on future technological breakthroughs and safety considerations. The authors acknowledge that the evidence points to rapid progress but stop short of claiming imminent self-sufficient AI evolution.
Next Steps in Monitoring AI Self-Development
Researchers and industry observers will likely focus on tracking further improvements in AI’s decision-making capabilities and its ability to autonomously design experiments. Future internal data releases and benchmark developments will help clarify whether the trend toward recursive self-improvement is continuing or accelerating. Regulatory and safety discussions are expected to intensify as AI systems take on more autonomous roles in research and development.
Key Questions
What is recursive self-improvement in AI?
Recursive self-improvement refers to AI systems autonomously improving their own capabilities, potentially leading to rapid, exponential progress without human intervention.
How does Anthropic’s data support claims of AI accelerating its own development?
Anthropic presents internal and public benchmark data showing AI models are increasingly handling tasks like coding and experimentation, with measurable growth over recent years, indicating faster internal progress.
Is full autonomous self-improvement currently happening?
No. While AI is automating many research tasks, the critical decision-making aspect—choosing which problems to pursue—remains human-controlled, and there is no confirmed evidence that AI can independently design its own successors yet.
Why does this matter for the future of AI?
If AI can fully automate its development process, it could lead to rapid advancements and new risks, making oversight and safety measures more urgent than ever.
Source: ThorstenMeyerAI.com
