📊 Full opportunity report: The Co-Founder’s Black Hole — A Structural Read on Jack Clark’s Automated AI R&D Essay on ThorstenMeyerAI.com — validation score, market gap, and execution plan.
TL;DR
Jack Clark, co-founder of Anthropic, forecasts a greater than 60% probability of AI systems autonomously conducting research without human involvement by 2028. This prediction is based on converging evidence from benchmarks and technical trends, highlighting a potential structural shift. The development raises urgent questions about institutional preparedness and AI safety.
Jack Clark, co-founder and head of policy at Anthropic, publicly forecasts a greater than 60% probability that AI systems will autonomously conduct research and develop successors without human intervention by the end of 2028. This is the first time a leading AI institution has made such a specific probabilistic prediction, signaling a potential paradigm shift in AI development and policy considerations.
On May 4, 2026, Clark published ‘Import AI #455’, where he states that there is over a 60% chance that AI systems capable of autonomously building their own successors will emerge within the next 32 months. The forecast is supported by a convergence of evidence, including six benchmarks showing rapid saturation in AI research capabilities, and exponential improvements in AI training speeds and performance metrics. Clark’s analysis suggests that the technical trajectory is approaching a threshold where recursive self-improvement could become feasible, raising profound questions about the control and safety of such systems.
The forecast is significant because it marks a shift from speculative warnings to institutional-level commitments, with Clark’s statement carrying weight within the AI research community. The forecast’s timeline aligns with key institutional milestones, such as Anthropic’s IPO and its post-IPO disclosures, implying that the industry and policymakers must prepare for a potentially transformative period within the next three years. However, the precise nature of how these developments will unfold remains uncertain, especially regarding the technical feasibility of fully autonomous research and the ability of current institutions to manage associated risks.
The black hole
is visible.
Four threads converge. One window. Anthropic’s head of policy has publicly committed to crossing a civilizational threshold within 32 months.
The structural feature of Clark’s argument is not that we cross a boundary and continue forward; it is that beyond a certain threshold, the forecastability of subsequent events degrades dramatically. We can see the geometry around the threshold. We can estimate when we will reach it. We cannot model what happens on the other side. The black hole event horizon analogy is precise.
Four pieces. One argument.
The four prior pieces in this series each addressed a single thread of Clark’s argument. The threads are independently significant. What this synthesis argues: they converge on a structural finding larger than any individual thread.
AI Workflow Automation for Bloggers: Build a Simple Content System to Research, Write, Optimize, and Repurpose Posts Faster with AI and No-Code Tools (AI Toolkit for Bloggers 2026 Book 8)
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Four threads. Four convergence arguments.
The threads converge structurally rather than independently. Each pair of threads produces a specific structural argument. The aggregate is larger than the parts.
The Coming Wave: AI, Power, and Our Future
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Clark’s essay doesn’t say.
Each sub-piece identified per-thread omissions. The synthesis level has its own omissions — features of the integrated argument that don’t appear in any single sub-piece but emerge when the threads are read together. Each is a real coordination problem with no resolution at scale.
NVIDIA Jetson Orin Nano Super Developer Kit
The NVIDIA Jetson Orin Nano Developer Kit sets a new standard for creating entry-level AI-powered robots, smart drones,…
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Thirty-two months. Five markers.
From May 4, 2026 to December 31, 2028 is 32 months. The trajectory either delivers the threshold Clark forecasts or it doesn’t. Specific indicators along the way that resolve the synthesis read in either direction.
- Clark publishes 60%/2028
- METR ~12 hr
- SWE-Bench 93.9%
- CORE solved
- Anthropic IPO prep
- METR ~100hr target
- SWE saturated
- MLE-Bench saturating
- PostTrain 40-50%
- Anthropic IPO Q4
- METR 300-500hr
- MLE saturated
- PostTrain at human
- RSI demo non-frontier
- 30%/2027 evidence
- METR 1K-3K hr
- “Trains successor” demos
- Alignment claims
- Catastrophic-risk window
- Stage 2 visible
- METR ~10K hr (naive)
- Automated AI R&D OR
- Inflection visible
- Machine economy Stage 3
- Black hole crossed
Autonomous Minds: How Agentic AI Predicts and Learns to Enable Productivity and Empowerment
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Five errors. Honest probabilities.
A serious analysis owes the reader an explicit account of where it could be wrong. Five categories of potential error in the synthesis above. The structural finding survives at lower forecast probabilities but is less acute.
Three parts. One window.
The four threads converge. The synthesis-level omissions sharpen the picture. The structural finding is the answer to “what does the Clark essay actually tell us, and what does it imply we should do?”
The black hole is visible. The event horizon is 32 months out. We can see the geometry around the singularity. We cannot see past it. What we can do during the window is build the institutional response that will determine what we encounter on the other side.
Implications for AI Policy and Safety in the Next 32 Months
This forecast underscores an urgent need for reevaluating AI safety protocols, regulatory frameworks, and institutional capacity. If Clark’s prediction proves accurate, the next 32 months could see the emergence of AI capable of independently advancing its own capabilities, potentially outpacing human oversight. Current institutional structures are not designed to handle such rapid, autonomous development, which could lead to unforeseen risks, including loss of control or unintended consequences. The forecast emphasizes that the window for proactive policy and safety measures is closing rapidly, making this period critical for shaping the future of AI governance.
Technical Trends Supporting the Autonomous AI Forecast
Multiple benchmarks and technical indicators support Clark’s forecast. Six different AI capability benchmarks—covering research speed, problem-solving, and fine-tuning—have shown exponential growth, with saturation points approaching thresholds necessary for autonomous research. For example, AI training speeds have increased 52-fold since 2025, surpassing human performance benchmarks by an order of magnitude. The trajectory of these improvements suggests that by late 2028, AI systems could reach a level where they can independently identify research problems, develop solutions, and iterate without human input. This convergence of technical progress indicates that the threshold for autonomous research may be imminent, aligning with Clark’s timeline.
Prior to this forecast, most predictions about AI takeoff were speculative or based on limited data. Clark’s institutional positioning and the convergence of multiple technical trends now provide a more concrete basis for assessing the likelihood of a near-term shift towards autonomous AI research.
“there’s a likely chance (60%+) that no-human-involved AI R&D — an AI system powerful enough that it could plausibly autonomously build its own successor — happens by the end of 2028.”
— Jack Clark
Uncertainties Surrounding Technical Feasibility and Institutional Readiness
While the technical trends and benchmark saturation support a high likelihood of autonomous AI research emerging by 2028, significant uncertainties remain. It is unclear whether current AI architectures can fully realize recursive self-improvement without unforeseen technical barriers. Additionally, institutional capacity to regulate, oversee, and respond to such systems is currently inadequate, but the exact timeline and nature of these gaps are still evolving. The possibility of technical or policy obstacles delaying or preventing this transition cannot be ruled out, and the model’s assumptions about exponential progress may not hold if fundamental challenges arise.
Next Steps for Monitoring and Policy Preparation
Researchers, policymakers, and industry leaders need to closely monitor the technical developments and benchmark saturation trends over the coming months. Key actions include developing safety protocols tailored to autonomous research systems, engaging in international policy discussions, and preparing institutions for rapid response scenarios. Further analysis is required to assess the technical feasibility of recursive self-improvement and to refine risk assessments. The period ahead will be critical for shaping the regulatory and safety frameworks necessary to manage the potential emergence of fully autonomous AI systems.
Key Questions
What does ‘autonomous AI research’ mean in this context?
It refers to AI systems capable of independently identifying research problems, developing solutions, and iterating on their own without human intervention, potentially leading to self-improving AI capable of building successors.
How reliable is Jack Clark’s forecast?
The forecast is based on converging technical indicators and institutional statements. However, uncertainties about technical feasibility and institutional capacity mean the prediction remains probabilistic, not certain.
Why is the 2028 timeline significant?
It marks a period within which the convergence of technical progress and institutional readiness could produce a transformative shift in AI capabilities, with profound policy and safety implications.
What are the risks if autonomous AI research is achieved?
Potential risks include loss of human control, unintended behaviors, and rapid, unpredictable technological advancement that current safety measures may be unprepared to handle.
What should institutions do now?
They should enhance safety protocols, develop international regulations, and prepare for rapid response scenarios to mitigate risks associated with autonomous AI systems emerging in the near future.
Source: ThorstenMeyerAI.com
