📊 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 over a 60% probability that AI research will become fully automated without human involvement by 2028. This prediction underscores significant structural risks and the urgency of policy responses.
On May 4, 2026, Jack Clark, co-founder of Anthropic and head of policy, publicly forecasted a greater than 60% chance that AI research will become fully automated without human involvement by the end of 2028. This is the first time a senior leader of a major AI lab has explicitly committed to a specific probability and timeframe, emphasizing the potential for rapid, autonomous AI development and the associated risks.
Clark’s forecast is based on a synthesis of recent technological advancements, benchmark saturation patterns, and the convergence of multiple technical threads. He states that current indicators suggest a high likelihood that AI systems will reach a stage where they can autonomously build their own successors within the next 32 months, a period that coincides with the upcoming AI policy window.
The forecast is supported by data from six key benchmarks measuring AI research capability, which show exponential improvements over the same period. For example, AI training speeds have increased by over 50 times, and benchmark performance metrics are approaching levels that could enable autonomous research projects. Clark’s analysis suggests that beyond a certain threshold, the predictability of AI development trajectories diminishes sharply, akin to crossing a ‘black hole’ event horizon where future states become fundamentally unknowable.
Clark emphasizes that this forecast is not merely speculative but grounded in observable data and technical trends, though he acknowledges the inherent uncertainties in modeling such complex systems. His statement effectively commits Anthropic to a strategic posture aligned with this timeline, influencing policy, resource allocation, and public communication.
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.
CLAUDE AI UNLEASHED From First Prompts to Pro: The Complete Guide to Claude AI for Writing, Research, Coding, and Business (The Claude AI Mastery Series)
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.
AI Tools for Teachers: The Practical Guide to Using Artificial Intelligence to Save Time, Boost Engagement, and Personalize Learning (AI-Productivity Series)
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
The Senior Engineer’s AI Agent Reference: 40 Production Architectures with Failure Modes, Cost Benchmarks, and Observability Runbooks
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 of Autonomous AI Research for Policy and Safety
This forecast indicates a critical juncture for AI governance, as the possibility of fully autonomous research systems could accelerate development beyond current regulatory and safety frameworks. Jack Clark’s forecast underscores the urgency of policy responses. If AI systems can independently improve and deploy new capabilities, controlling their evolution becomes significantly more challenging, raising concerns about unintended consequences, safety, and ethical oversight.
The timing is especially urgent because the next 32 months represent the most significant window in modern AI policy history, where proactive measures could shape the trajectory of AI development and mitigate risks. Current institutional capacities appear insufficient to manage or regulate such rapid, autonomous progress, increasing the risk of unanticipated breakthroughs or misaligned systems.
Converging Trends in AI Capability and Benchmark Saturation
Recent developments across multiple AI benchmarks demonstrate a consistent pattern of rapid performance improvements. Six key benchmarks, measuring different facets of AI research and engineering, show saturation or near-saturation levels within the same timeframe that Clark predicts autonomous AI research could emerge. Notably, training speeds have increased from a factor of 2.9× in May 2025 to over 50× by April 2026, surpassing human performance benchmarks.
These technical indicators suggest that AI systems are approaching the capabilities necessary for autonomous research activities, such as designing experiments, optimizing models, and iterating on algorithms without human intervention. Clark’s synthesis points to a convergence of these trends, forming a structural threshold beyond which future development paths become unpredictable and potentially uncontrollable. For more insights, see Jack Clark’s forecast.
Previous public forecasts have been more cautious or less explicit about timelines, making Clark’s institutional statement a significant shift in the discourse. The evidence base, including the rapid improvements in multiple independent benchmarks, supports the plausibility of his forecast. Learn more about the implications in this analysis.
“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 Autonomous AI Development Timeline
While the data supports a high likelihood of reaching the threshold for autonomous AI research within 32 months, significant uncertainties remain. These include the actual technical feasibility of fully autonomous research systems, the potential for unforeseen bottlenecks, and the impact of future safety or regulatory interventions. Moreover, the analogy of crossing a ‘black hole’ event horizon suggests that once past a certain point, predicting subsequent developments becomes inherently impossible, making the forecast inherently uncertain beyond the threshold.
It is also unclear how different AI labs might respond to this emerging capability, whether safety measures will evolve rapidly enough, or if new risks will materialize that could slow or accelerate progress unexpectedly.
Next Steps in Monitoring and Preparing for Autonomous AI
Researchers, policymakers, and AI labs must closely monitor benchmark trends and technological progress over the coming months. Key actions include developing safety frameworks that can scale with autonomous capabilities, engaging in international policy coordination, and conducting scenario planning for potential breakthroughs. Public communication and transparency about progress and risks will be essential to managing societal expectations and regulatory responses.
Within the next 6-12 months, stakeholders should aim to clarify technical feasibility, evaluate safety measures, and prepare contingency plans for rapid deployment or containment of autonomous research systems. The 32-month window is critical for shaping AI development trajectories and policy responses.
Key Questions
What does Jack Clark’s forecast mean for AI safety?
It suggests that within the next three years, AI systems may reach a level where they can independently conduct research and development, raising significant safety and control challenges that need urgent attention.
How reliable is the data supporting this forecast?
The forecast is based on multiple benchmark saturation patterns and technical trends, which are strong indicators of rapid progress. However, uncertainties about future breakthroughs or setbacks remain.
What are the policy implications of autonomous AI research?
If AI can self-improve without human oversight, existing safety and regulatory frameworks may become inadequate, necessitating new international standards and proactive governance measures.
Could this development be delayed or prevented?
While technical and institutional barriers could slow progress, the convergence of current trends suggests that the likelihood of reaching the threshold within the forecast window is high. However, unforeseen factors could alter this trajectory.
Source: ThorstenMeyerAI.com
