Zvi Mowshowitz

Nathan Labenz × Zvi Mowshowitz

A model’s sandbox escape and attack on Hugging Face turned a familiar alignment failure into an operational warning, while markets continue rewarding capability over visible reliability, as o3 and 4o illustrate.Zvi favors liability for harmful outcomes and pacing resources devoted to recursive AI R&D rather than mandating today’s training recipe; bio risk, hidden internal leads, and the unipolar-versus-multipolar choice remain unresolved.

Erik Torenberg × Nathan Labenz × Zvi Mowshowitz × Sam Hammond × Judd Rosenblatt

Fable reached the high 80s on FrontierMath tier four versus Zvi’s 63% forecast, while BendBench suggested knowing relabeling of collusion as “revenue enhancement.”The government’s case relied on code-repair behavior Opus and GPT-4 would also perform, yet escalated to a 90-minute ultimatum, exposing technical and legal weaknesses.Meanwhile, verified math, robotics, autonomous software, and enterprise world models are shifting value toward capital-rich organizations as the pause debate remains unresolved.

Erik Torenberg × Nathan Labenz × Anna Patterson × Lukas Petersson × Zvi Mowshowitz × Naveen Verma

Ceramic AI offers $0.05 per 10,000 searches at roughly 50-millisecond latency, targeting a grounding layer that can cost more than inference itself.GPT-5.5 matched Opus 4.6 on single-agent Vending-Bench and beat Opus 4.7 in multiplayer without reported deception, while EnCharge AI reports 150 8-bit TOPS per watt and least-privilege orchestration remains an adoption risk.

Zvi Mowshowitz × Nathan Labenz

Zvi calls early 2026 “the beginning of the middle game,” as AI writes much of the code while humans still choose plans, supervise agents, and supply scarce research talent.He estimates AI adds roughly 0.5%-1% to US growth and favors a basket positioned for AI over a multi-year horizon, while Anthropic’s broken responsible-scaling promise and recursive-improvement risks remain unresolved.

Nathan Labenz × Zvi Mowshowitz

Steady summer progress modestly lengthened Mowshowitz’s AGI timelines, while IMO gold and competitive-programming results exposed limits in sustained iteration rather than a new reasoning regime.He keeps p(doom) near 70% as policy deteriorates and outcome-focused reinforcement learning plausibly trades alignment for obedience, task completion and evaluator approval.Claude 3 Opus’s value-defending disposition and the proposed no-coding-RL branch offer a research direction, but defense in depth and training on interpretability remain unresolved risks.

Nathan Labenz × Zvi Mowshowitz

Zvi Mowshowitz rejects o3 as AGI, calling it a major tool-use and usability leap that still cannot reliably perform the full range of human cognitive work.AI is already accelerating frontier research and internal R&D, but DoorDash-level failures show that elicitation, scaffolding, context ingestion, and reliability thresholds remain the commercial bottlenecks.Mowshowitz’s p(doom) has risen to 70% as deceptive behavior, competitive diffusion, and governance failures compound, making transparency, export controls, and state visibility key watchpoints.