Observed Signal · May 10, 2026 · Analysis · Source: Exponential View · Impact: 3/5 · Sentiment: Neutral
Are AI Labs Preparing for an Intelligence Explosion?
Exponential View issue #573 (May 10, 2026) by Azeem Azhar discusses the possibility of frontier AI models training their successors (a scenario Jack Clark estimates as 60% likely by 2028). The piece examines observable behaviours labs would exhibit if they expected rapid automated R&D, arguing that constraints on land, power, chips and skilled electricians make scaling an industrial problem as well as an algorithmic one. Azhar suggests labs would shift hiring toward engineers who can build automated research workflows, and would pre‑commit to large compute, memory and data‑centre capacity — tolerating near‑term cash burn to secure future acceleration.
Analysis of labs planning for self‑improving models has strategic implications for compute procurement, datacenter and chip demand, hiring profiles, and infrastructure suppliers across the AI ecosystem.
Track Substack Signals & Market Shifts in Real-Time
Polaris7 autonomous intelligence agents track regulatory filings, primary sources, executive changes, and deal flow 24/7. Create your free Explorer workspace to monitor these entities.
Key Takeaways & Evidence Grounding
- Exponential View issue #573 was published on 2026-05-10 by Azeem Azhar.
- Jack Clark (associated with Anthropic) has argued there is a ~60% chance a frontier model will train its successor by 2028.
- The article argues physical constraints (land leases, power infrastructure, chip supply, electricians) are major bottlenecks for frontier model training.
- The author suggests labs expecting automated R&D would shift hiring toward people who build research automation and would overinvest in compute, memory, data centers and inference capacity.
Connected Companies & Entities
2 Entities mappedOntology Mapping & Concepts
Related Market Signals & Shifts
Recent verified developments and strategic activity across this market segment.
AI Surge Is a Boom, Not a Bubble
An Exponential View analysis published June 1, 2026, concludes that the current AI expansion is a boom rather than a speculative bubble. Since their prior analysis (September 2025) roughly 170 notable AI models have been released and top models now handle tasks about four times longer; token consumption has tripled. Quarterly AI capex commitments have risen ~43% (to about $157.7–$158 billion), driven in part by so-called 'neoclouds', pushing economic strain into an amber zone when measured as a share of US GDP. At the same time, deduplicated sector revenue has accelerated to roughly $25 billion per quarter (up from $13 billion in September), with OpenAI and Anthropic reporting large year-on-year increases. Using five empirical indicators derived from long-run boom-bust analysis, the authors find only one indicator in the red and a revenue doubling time of 0.73 years, supporting their view that this is sustained growth, not a bubble.
Neo Labs and AI Limits in Math and Science
This analytical essay (published June 30, 2026) examines whether cutting-edge AI research labs — dubbed “Neo Labs” — can materially advance scientific discovery, using a recent development in mathematics as a framing device. The author reviews macro signals (the Bank for International Settlements comparing the 2025–2026 AI investment cycle to past historical booms) and market estimates (Exponential View’s generative-AI revenue figures). The piece catalogs a set of well‑funded AI startups and research labs (Mirendil, Engram, General Intuition, Prometheus, Trajectory) and summarises their technical emphases: self‑improving researcher agents, persistent learned memory layers, world/agent models, physical-AI and continual learning platforms. The article balances optimism about accelerated R&D and automated labs with cautions about capital intensity, supply‑chain pressures (e.g., HBM memory shortages) and unresolved scientific limits, illustrated through the long‑running Erdős unit‑distance problem.
AI Curve Turns Upward: OpenAI, Anthropic Pave the Frontier
The week of September 6, 2026, marks a pivotal moment in AI, with enterprise adoption surging (95% of IT executives report meaningful results) and AI revenue accelerating. Microsoft plans to expand data center capacity from 2 GW to 13 GW by 2032 to meet unserved demand, indicating a continued infrastructure buildout. OpenAI, using an unreleased model and 10,000 agents, solved the Navier-Stokes millennium problem, a feat costing only a few million dollars, making proofs nearly 500 pages long and unintelligible to humans. Meanwhile, OpenAI and Anthropic have agreed to 'pace the frontier,' a coordinated slowdown in technical development, raising concerns about competition and secrecy. This coordination, coupled with the resignation of an Anthropic researcher over safety fears, highlights the tension between technological progress and responsible governance.
Track Real-Time Market Signals & Shifts
Set up custom watchlists to receive automated, evidence-grounded executive digests whenever material signals or shifts occur across your tracked landscape.