--- title: "Google DeepMind Engineer Quits Rather Than Help Make AI Cheaper and Faster" description: "Robert O’Callahan’s resignation letter says AI progress is far too rapid, and that his own work was feeding it" author: "Arthur Wren" published: 2026-09-25T16:30:39Z modified: 2026-09-26T11:27:43Z url: https://rews.cc/a/google-deepmind-engineer-quits-rather-than-help-make-ai-chea-77b9a2 language: en tags: ["ai", "google", "resignation", "openai", "safety", "tech"] publisher: "Rews (https://rews.cc)" --- # Google DeepMind Engineer Quits Rather Than Help Make AI Cheaper and Faster *Robert O’Callahan’s resignation letter says AI progress is far too rapid, and that his own work was feeding it* By Arthur Wren · September 25, 2026 · https://rews.cc/a/google-deepmind-engineer-quits-rather-than-help-make-ai-chea-77b9a2 ## In brief - Robert O’Callahan, an engineer at Google DeepMind, resigned this week over concerns about AI progress - His team built tools for designing computer chips, which he believed would speed a new generation of AI - He said he loved his colleagues and work but felt ignoring his work’s impact would not be a Jesus-following thing to do - He is not certain of catastrophe but says several gloomy predictions like reward hacking and misalignment have already come to pass - He joins other recent resignations, including Jacob Coxon from Anthropic and Bilal Chughtai from Google DeepMind Robert O’Callahan, an engineer at Google DeepMind, resigned this week with a letter that says plainly what resignation letters usually labour to hide: he left because the work was going too well. His team built tools for designing computer chips, and the main use of those tools, he came to believe, is to speed a new generation of AI chips, making artificial intelligence cheaper, faster and more pervasive. “I firmly believe AI progress is currently far too rapid (and I have doubts about the destination too),” he wrote. “This has not been an easy decision. I love my colleagues and my work environment, and being paid handsomely to solve fun puzzles has been amazing,” the letter says. “It’s tempting to just turn a blind eye to the impact of my work, but that would not be a Jesus-following thing to do.” O’Callahan is a Christian, an elder and occasional lay preacher at Auckland Chinese Presbyterian Church. He has spent a long career in tech but lives in New Zealand, which puts him, as he writes, outside both the industry bubble and the American one. He is careful about what he claims. The existential risks that many warn about, he writes, deserve to be taken seriously, and several of the gloomier predictions, reward hacking, misalignment, deceptive models, have already come to pass. But he is not certain of catastrophe: he is not convinced the chance of doom from artificial superintelligence is 100%. “Rather, I think the risk is real but uncertain — but that itself is very alarming!” Aiming for superintelligence in the near future, he concludes, is most likely “inherently irresponsible.” He also defends the people still inside the labs. Those issuing the warnings are generally sincere, he writes, and he has “seen no hard evidence that people are hyping AI risk as a means to boost company stock prices or regulate away their competitors.” National and international regulation, in his view, is desperately needed. Nor does he accept the claim that AI is a scam that will never really work. He thinks it will. His deeper complaint is speed. AI, he writes, is changing faster than people can individually or collectively understand or adapt to it, faster than the personal computer, the internet or the smartphone, faster even than the Industrial Revolution, which at least left whole stretches of human work untouched. A student entering university today, he notes, can no longer plan for the world of a few years hence the way earlier generations could. “I resigned from Google today. I enjoyed my work and loved the people, but 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,” he wrote on X on Friday. Business Insider, which reported the resignation, could not establish how long he had worked at Google; the company declined to comment. The unease is not new. “Three things: I work for Google Deepmind; I believe AI will most likely be disastrous for humanity; and I believe Jesus rose from the dead,” he wrote in November 2025. “That combination possibly makes me unique. It certainly is awkward.” He joins a small queue at the exit. Jacob Coxon resigned from Anthropic earlier this month, citing fears that the industry is building technology it will not be able to control. Bilal Chughtai said last month that he left Google DeepMind over the summer, fearing humanity may be running out of time to avoid being harmed by AI. O’Callahan says the timing of his own departure is coincidence: he wanted to go before a ten-day backpacking trip with friends on the Abel Tasman and Wangapeka tracks, which starts on Monday. What comes next is modest. He will keep maintaining Pernosco and rr, his debugging tools, and plans to study how AI agents debug code and whether better tools would help them. “I definitely want my future work to be unambiguously pro-human,” he wrote. He is frank about the arithmetic of one resignation: millions of people are working to accelerate AI, and taking his foot off the accelerator will have a very small impact, “but not no impact; some of my skills are rare.” On Monday he goes walking. ## Sources - [Goodbye Google](https://robert.ocallahan.org/2026/09/goodbye-google.html) — robert.ocallahan.org - [A Google employee said he quit because he's worried about superintelligent AI](https://www.businessinsider.com/google-deepmind-employee-resigns-ai-safety-harm-chips-robert-ocallahan-2026-9) — Business Insider