Peter Diamandis
Episode overview

Researcher Quits: “Gambling With Our Lives,” Altman Calls for Slowdown, and Labs Race

Peter Diamandis · 11 September 2026 · 2h 43m · 1 Egleze moment
Researcher Quits: “Gambling With Our Lives,” Altman Calls for Slowdown, and Labs Race ▶

What happened in this episode?

Peter Diamandis and a panel including Emad Mostaque, Salim Ismail and other guests debate AI acceleration, safety messaging and the economics of scaling. A central segment focuses on a viral resignation post attributed to Jacob Conix, described as an Anthropic researcher, and a separate claim from an Anthropic alignment lead that the chance of “kill all humans” outcomes exceeds 10% in a decade; one panellist says they discount such exits as performative but alleges the post’s reach could indicate a foreign influence operation. The group also discusses Sam Altman’s reported call to “pace progress” after OpenAI’s claimed Navier–Stokes breakthrough and broader expectations that AI will rapidly solve major maths and science problems.

On governance, Diamandis cites an FT report that Anthropic declined to provide a new model to the UK AI Security Institute for pre-release testing, which the panel frames as evidence that model weights and evaluation access are becoming national-security sensitive. In geopolitics and competition, Diamandis cites Bloomberg reporting that ByteDance founder Zhang Yiming is personally overseeing a real-time spatial “world model”.

The episode covers market signals of compute scarcity, including a claimed 22% month-on-month rise in H100 rental prices, and discusses Chinese model releases and memory-efficiency techniques as pressure on Western inference economics. In health and longevity, Diamandis highlights Insilico Medicine’s AI-designed drug rentosertib moving into Phase 3 for idiopathic pulmonary fibrosis and cites Phase 2a biomarker analysis suggesting multi-clock shifts consistent with reduced biological age. The panel also highlights DeepMind’s Alpha Genome Atlas as a precomputed map of variant effects across billions of potential single-letter genome mutations.

What are the key points?

Who is speaking?

Named speakers identified in Egleze's source-linked moments from this episode.

Which topics are covered?

How is this page sourced?

Programme
Peter Diamandis
Published
11 September 2026
Duration
2h 43m
Original episode
Open the original source ↗

Egleze verifies the speaker, source and timestamp for each linked moment. Unless a moment is explicitly marked independently verified, statements remain attributed claims by the named speaker rather than established facts.

Watch original episode More from Peter Diamandis

1 moments from this episode

Source-linked · editorially selected
01
AI & Tech

Diamandis reads Altman-attributed statement urging slower AI progress after claim

Peter Diamandis (reading a statement he attributes to Sam Altman) · at 48:01 · Attributed speaker claim
The world has extremely capable models now. I did not expect a result of this magnitude to happen so soon. We've been talking a lot about the need to pace progress to ensure safety. For me, this is the strongest evidence yet for that urgency.

Peter Diamandis reads a statement he attributes to OpenAI chief executive Sam Altman, linking OpenAI’s claimed Navier–Stokes breakthrough to calls to slow AI progress for safety. Diamandis then invites discussion about whether the messaging reflects genuine concern, marketing, or politics.