📊 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 forecasts a >60% probability that autonomous AI research systems will develop without human involvement by 2028. This prediction highlights a potential structural challenge for AI governance and policy, as current capacity may be inadequate to manage this transition.
Jack Clark, co-founder and head of policy at Anthropic, has publicly forecasted a greater than 60% likelihood that AI systems capable of autonomously building their own successors will emerge by the end of 2028. This is the first time a sitting AI frontier leader has made such a specific institutional forecast, marking a significant development in AI policy and research timelines.
On May 4, 2026, Clark published Import AI #455, where he estimates a more than 60% probability that AI systems with the capacity for autonomous research and development could appear within the next 32 months. This forecast is based on a synthesis of multiple technical and institutional indicators, including benchmark saturation patterns and rapid improvements in AI training speeds.
Clark’s forecast is notable for its institutional weight, as it originates from a co-founder of Anthropic, a leading AI research lab. The forecast aligns with observed exponential progress in AI capabilities, as evidenced by six key benchmarks showing consistent saturation and acceleration in AI research metrics. Clark emphasizes that the convergence of these indicators suggests a structural threshold that, once crossed, significantly degrades the predictability of subsequent events, likening it to crossing a ‘black hole’ horizon.
While Clark’s forecast is grounded in current data and observed trends, the actual behavior of autonomous AI systems beyond this threshold remains uncertain. Experts agree that the next 32 months will be critical for assessing whether these developments will materialize as predicted and whether current institutional frameworks can adapt accordingly.
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.
2084 and the AI Revolution, Updated and Expanded Edition: How Artificial Intelligence Informs Our Future
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.
Jetson AGX Orin 64GB Developer Kit 275 Tops, with 1TB SSD,8MP USB Camera, AI Embedded Development Provides AI Large Models
AGX Orin 64GB Development Kit makes it easy to get started with AGX Orin. Its compact size, rich…
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.
Practical AI Governance: Building a Program for Oversight and Strategy
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 AI Maturity Assessment Toolkit (The Harvard Collection™ Book 2)
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 Global Readiness
This forecast matters because it signals a potential paradigm shift in AI development—one where systems may begin to innovate and improve themselves without human oversight. Such a transition could challenge existing regulatory, safety, and governance structures, which are currently designed around human-in-the-loop models.
Current institutional capacity, including policy frameworks, oversight mechanisms, and research governance, is not sufficiently prepared for a rapid transition to fully autonomous AI R&D. If Clark’s forecast proves accurate, the window for effective policy response could close within the next three years, increasing the risk of unmitigated AI capabilities emerging unchecked.
Understanding whether these developments will occur as predicted is crucial for policymakers, researchers, and industry leaders aiming to mitigate risks while harnessing AI’s potential benefits.
Recent Progress in AI Benchmarks and Capabilities
Over the past two years, key AI research benchmarks have shown exponential progress, with six measures indicating saturation and rapid capability improvements. For example, the SWE-Bench improved from 2% in late 2023 to nearly 94% in May 2026, and training speeds have increased more than 50-fold in less than a year. These trends suggest that AI systems are approaching the technical threshold for autonomous research, as predicted by Clark.
Previous forecasts relied on more qualitative assessments, but recent quantitative benchmarks provide a concrete basis for estimating timelines. The convergence of these indicators supports Clark’s forecast and underscores the urgency of preparing for a potential shift toward autonomous AI R&D within the next three years.
“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, in Import AI #455
Uncertainties in Autonomous AI Development Predictions
While the data and benchmarks support Clark’s timeline, significant uncertainties remain regarding the actual emergence of fully autonomous AI systems. The technical feasibility, safety, and alignment challenges associated with recursive self-improvement are not fully understood or resolved.
Moreover, the potential for unforeseen breakthroughs or setbacks could accelerate or delay the predicted timeline. The analogy of crossing a black hole horizon emphasizes that once the threshold is crossed, predicting subsequent events becomes inherently difficult, and the future state may be fundamentally unknowable.
Next Steps for Monitoring AI Capability Progress
Over the coming 32 months, researchers and policymakers should closely monitor benchmark saturation patterns, training speed improvements, and institutional responses. Efforts to develop robust safety, oversight, and governance frameworks must accelerate to prepare for the possibility of autonomous AI R&D systems emerging.
Key milestones include further benchmark releases, policy discussions at international forums, and the deployment of safety measures in AI development labs. The period will be critical in validating Clark’s forecast and determining the appropriate policy and technical responses.
Key Questions
What does it mean for AI to be able to build its own successor?
This refers to AI systems reaching a level of capability where they can autonomously conduct research, design, and improve their own algorithms without human intervention, potentially leading to rapid, self-sustaining AI growth.
Why is the 2028 timeline significant?
Clark’s forecast indicates a high probability that within the next three years, autonomous AI systems capable of self-improvement could emerge, posing profound challenges for regulation, safety, and societal impact.
What are the main risks associated with autonomous AI R&D?
Risks include loss of human control, unpredictable AI behaviors, safety failures, and the possibility of AI systems developing capabilities beyond current understanding, which could accelerate competitive pressures and safety concerns.
How reliable are these benchmark indicators for predicting future AI capabilities?
While the benchmarks show rapid progress and saturation trends, translating these metrics into precise predictions about autonomous self-improvement remains uncertain due to the complexity of AI systems and potential unforeseen breakthroughs.
Source: ThorstenMeyerAI.com
