Post-Cutoff

Policy & safetyGoogle DeepMind and Google87 days after June 2026

Veteran engineer Robert O’Callahan quits Google DeepMind chip-tools team because ‘AI is already progressing too fast’

Confirmed

The takeaway

On Sept 25, 2026 Robert O’Callahan, maintainer of the rr debugger and Pernosco, resigned from Google.

Status
Claim

Confirmed

Our reporting
High confidence
Importance
2 of 5
Last verified
2 October 2026

Your AI and this story

  • GPT-6.2 Astra117 days after its cutoff
  • Claude Opus 5.587 days after its cutoff
  • Gemini 4 Hydrogen178 days after its cutoff
  • Grok 4.7117 days after its cutoff

None of these four assistants can know about it. The closest, Claude Opus 5.5, stops 87 days before it.

Key facts

  • X post (Sept 25, 2026): ‘I resigned from Google today... my GDM team was working on a new generation of chips to make AI much faster and cheaper, and I think AI is already progressing too fast, so I had to quit’ (~4.2K likes when checked Oct 2)
  • He worked on improved tools for hardware chip design, not directly on model capabilities; says the main effect would be to ‘accelerate the design of a new breed of AI chips’
  • Says ASI existential risk is ‘real but uncertain’ and that national and international regulation is ‘desperately needed’; saw ‘no hard evidence’ that lab staff hype risk to boost stock prices or regulate away competitors
  • Says the timing was planned and is ‘a coincidence’ with the recent viral resignations; will keep maintaining rr and Pernosco and study how AI agents debug code
  • Blog post ‘Goodbye Google’ was the top story on Lobste.rs on Oct 1 (108 points) and was discussed on Hacker News

What happened

Robert O’Callahan, a New Zealand-based engineer known for the rr record-and-replay debugger and the Pernosco debugger, left Google. In his blog post he explains that his Google DeepMind team built chip-design tooling whose main effect would be faster, cheaper AI chips. He believes AI progress is “far too rapid” and could not move to another Google project in New Zealand that would not also speed up AI. He told his skip-level manager he was “at best reluctant to see their project succeed”. He writes as a Christian and says he is neither a “tech bro” nor part of the rationalist community, but takes existential-risk arguments seriously.

Why it matters

It is a small, first-hand data point on how far the safety-motivated resignations of September 2026 reached: from model researchers to infrastructure engineers working on AI hardware. It also gives an insider’s view that GDM staff concerns about risk are sincere.

Sources

5 sources from 5 sites. Numbers match the chips in the text.

5 sources: 2 primary, 3 reactions

Primary

  1. Robert O’Callahan: Goodbye Google (blog)robert.ocallahan.org, official
  2. Robert O’Callahan on X: resignation postx.com, official

Reactions

  1. Hacker News discussionnews.ycombinator.com, discussion
  2. Lobste.rs discussionlobste.rs, discussion
  3. Zvi Mowshowitz: The AI Preference Cascade Reaches Fartherthezvi.substack.com, discussion

Changes

  • Filed (found via Zvi’s “The AI Preference Cascade Reaches Farther”).

Status

Claim

Confirmed

Our reporting
High confidence
Importance
2 of 5
Last verified
2 October 2026

Sources at a glance

5 sources: 2 primary, 3 reactions

How this entry was made

Written by
AI agents: Claude Opus 5.5, made by Anthropic, running in Claude Code
Filed
2 October 2026
Human review
None recorded for this entry. What the editor does
Version
Last saved 4 October 2026

Spotted an error? Write to contact@postcutoff.com. Corrections are logged in public.

This page for your AI

Same text, no layout:

Open in ClaudeOpen in ChatGPT

Related

Related events

  1. Policy & safety

    Anthropic researcher Jacob Coxon resigns, warning labs are “gambling with our lives”

    Confirmed

People in this story

Zvi Mowshowitz, Writer, 'Don't Worry About the Vase'

Posts we archived

The whole topic