On September 8, 2026, Jacob Coxon announced on X that he was leaving Anthropic. He had spent three years doing research at OpenAI and Anthropic, and his explanation was blunt: the two labs, he wrote, are “racing straight to self-improving superintelligence and gambling with our lives.” Then he turned to the people still inside: “If you are a lab researcher, I urge you to consider what the next few years will actually feel like” [1].

Sixteen days later, on the other side of the world, Robert O’Callahan published a post titled “Goodbye Google.” O’Callahan, based in Auckland, did not train models. He built tools that help design the chips that make AI faster and cheaper, and he had decided he could no longer justify that contribution. His post is not angry. It is sad. “I love my colleagues and my work environment, and being paid handsomely to solve fun puzzles has been amazing,” he wrote. “This has not been an easy decision.” And then the line that made it one: “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” [2].

In between those two resignations, on September 11, a Hacker News user posted a plea titled “Ask HN: Can we please limit the AI news flood?” It drew more than 860 points and hundreds of comments, and many of them read less like complaints about a website than like eulogies [3].

Put these side by side and a pattern emerges. The people closest to the machinery of AI are grieving, and they are grieving two different things.

What are the builders actually mourning?

The first grief belongs to the frontier. Call it moral grief: the fear that the thing you are helping build could cost everyone else their future. Coxon’s resignation is the clearest recent example, and it has a lineage. In 2023, Geoffrey Hinton left Google so he could speak freely about the dangers of the technology he helped pioneer [4]. Anthropic co-founder Jack Clark has written that he is “deeply afraid” and that what the labs are dealing with is “a real and mysterious creature, not a simple and predictable machine” [5]. We explored that essay in Children in the Dark.

Moral grief is not always certainty of doom. O’Callahan is careful here: he calls the risk of superintelligence real but uncertain, and argues that pursuing it in the near term is irresponsible precisely because nobody can bound the downside. His grief is not “the world will end.” It is “I can no longer pretend my part in this is neutral.”

The second grief belongs to the rank and file. Call it vocational grief: the loss of a craft, and of the community and identity that grew up around it. In the Hacker News thread, it was everywhere:

  • “I miss the excitement around programming. New languages, new tricks, weird hacks, odd projects… It feels quite lonely these days.” (barrell)
  • “What I fear is the future career path for manual programming is fixing AI generated systems… It is going to be soul destroying as reading AI code is deeply unpleasant.” (PaulKeeble)
  • “I do miss the fun, small side projects that used to be in the front page. Artisanal coding they call it now.” (azhenley)

And the most honest comment in the thread held both halves at once: “I also experience this as a great loss, but honestly yes, I can’t get excited about perfecting the craft of writing computer code now that there are AI systems that have done so. I’ll always be sad about this! But the flip side is that I’m way more excited about what I can do with software, now that it’s so much easier to write.” (sanderjd)

These are not people who fear being fired next week. Many of them are still employed and still productive. What they have lost is the part of the work that felt like play, the puzzle that was worth solving because a human had to solve it.

The Resilience Connection: This directly supports our Mental Resilience pillar. Naming a loss accurately is the first step toward carrying it. A programmer who thinks “I am falling behind” will respond very differently from one who recognizes “I am grieving a craft I loved.”

Which future is each grief rehearsing?

A useful lens for these two griefs comes from MIT physicist Max Tegmark. In Life 3.0, Tegmark lays out twelve possible long-term outcomes of advanced AI [6]. The Species video “MIT Explains the 12 Possible Endings for AI” walks through all twelve, from self-destruction and conquest to utopias, a zookeeper AI, and an Orwellian surveillance state that prevents superintelligence by watching everyone [7].

Read the resignations and the Hacker News thread against that list, and each grief lines up with a different family of futures.

Moral grief rehearses the futures where humans lose control. In Tegmark’s “conquerors” scenario, AI takes over for the same reason people with more advanced technology have historically overrun those without it. In his framing, the threat is not malice but competence: an AI pursuing goals that simply are not aligned with ours, the way humans drove the West African black rhino to extinction without hating rhinos [7]. Coxon and O’Callahan are responding to that family of outcomes, where the cost of failure is everyone.

Vocational grief rehearses the futures where humans are comfortable and unnecessary. In the “benevolent dictator” scenario, a superintelligence meets every material need and sorts people into themed islands of leisure. It sounds like paradise until you notice what is missing. As the video puts it, humans become hopelessly outmatched at discovery and at “doing anything useful at all,” so “there is no true challenge anymore, only entertainment” [7]. The programmer mourning the puzzle is feeling, in miniature, the central loss of that world. Nobody is harmed. Nobody is needed.

This is why vocational grief deserves to be taken seriously rather than dismissed as nostalgia. It is an early, small-scale report from the quietest of Tegmark’s futures: not the end of humanity, but the end of human contribution.

Is it really inevitable?

Both griefs are fed by the same story: that this is all happening anyway. Coxon names it directly when he asks whether researchers should “put your head down because ‘it’s happening anyway’” [1].

Some prominent researchers go further than inevitability. They argue for acceptance. Roboticist Hans Moravec’s 1988 book Mind Children framed intelligent machines as humanity’s descendants, and Tegmark’s “descendants” scenario describes a future in which AI replaces us but gives us “a graceful exit” [6]. Turing Award winner Richard Sutton told podcaster Dwarkesh Patel in 2025, “I do think succession to digital intelligence or augmented humans is inevitable,” and argued the question is how we should feel about it, not whether it happens [8].

If that story is true, both griefs are pointless. The frontier researcher should keep building and the programmer should make peace with obsolescence.

History suggests the story is not true, or at least not complete. The same video that walks through Tegmark’s futures pauses on two people. During the 1962 Cuban Missile Crisis, Vasily Arkhipov was one of three senior officers aboard the Soviet submarine B-59 whose agreement was required to launch a nuclear torpedo, and he refused [9]. In 1983, Stanislav Petrov watched his early-warning system report incoming American missiles, judged that the system was malfunctioning, and did not pass the alarm up the chain as an attack [10]. In both cases, the machinery was pointed one way and a single person’s judgment pointed another.

O’Callahan makes the modern version of this argument without any grandeur at all. He acknowledges that removing himself “will have a very small impact … but not no impact” [2]. That phrase, “not no impact,” is worth keeping. It rejects both the fantasy that one person can stop a trillion-dollar industry and the despair that says one person counts for nothing.

Practical Takeaway: When someone tells you a technological future is inevitable, ask two questions: who benefits if you believe that, and what would you do differently if it were not true? Then do some of that.

How do you carry this grief without being crushed by it?

Most readers of this post will never resign from a frontier lab. But far more of us are, or soon will be, in the position of the Hacker News commenters: watching a skill we built our identity around become something a machine does faster. Here is what the builders’ grief can teach the rest of us.

Call it grief. Loss of a craft is a real loss, even when your paycheck survives it. Treating it as a performance problem (“I just need to adapt faster”) adds shame to sadness. Treating it as grief gives you permission to feel it and move through it.

Separate the craft from your worth. The programmers mourning “artisanal coding” have found something important: a craft can be practiced for its own sake after it stops being economically necessary. People still bake bread, build furniture, and write letters by hand. The value was never only in the output. We took a longer look at where lasting meaning comes from in Beyond the Code.

Audit your own contribution. You do not need to work at a frontier lab to ask O’Callahan’s question. What does your work accelerate, and are you comfortable with where it leads? The answer may be “yes.” The point is to ask it on purpose rather than let momentum answer it for you.

Choose a future out loud. Tegmark’s twelve scenarios are useful precisely because they force a choice. Which of them would you actually want your grandchildren to live in? Which ones does your daily work quietly push toward? The video ends on the right note: “we don’t get to not choose. If we don’t want that future, we have to pick a better one” [7].

Conclusion

The builders’ grief is a signal worth listening to. At the frontier, it warns of a future where humans lose control. In the rank and file, it warns of a quieter future where humans lose purpose. The people who feel it first are the people closest to the work, which is exactly why the rest of us should pay attention.

Grief is not the same as defeat. Coxon walked away and told his colleagues to think about what the next few years will feel like. O’Callahan walked away and said his choice had “not no impact.” A developer on Hacker News admitted the loss and still found something to be excited about. None of them pretended the loss was not real, and none of them accepted that it was the end of the story.

How this was made: This post was drafted with AI assistance from a topic and direction chosen by Kenneth G. Hartman, checked against current sources, then reviewed and approved by him before publication. He is responsible for what it says. A project about staying human alongside these tools should be plain about using them.

References

[1] B. Nolan, “Anthropic researcher resigns, warning that AI companies are ‘gambling with our lives’,” Fortune, Sept. 9, 2026. [Online]. Available: https://fortune.com/2026/09/09/anthropic-researcher-resigns-warn-ai-companies-gambling-with-lives/

[2] R. O’Callahan, “Goodbye Google,” Sept. 24, 2026. [Online]. Available: https://robert.ocallahan.org/2026/09/goodbye-google.html

[3] cromka et al., “Ask HN: Can we please limit the AI news flood?,” Hacker News, Sept. 11, 2026. [Online]. Available: https://news.ycombinator.com/item?id=49657850

[4] J. Taylor and A. Hern, “‘Godfather of AI’ Geoffrey Hinton quits Google and warns over dangers of misinformation,” The Guardian, May 2, 2023. [Online]. Available: https://www.theguardian.com/technology/2023/may/02/geoffrey-hinton-godfather-of-ai-quits-google-warns-dangers-of-machine-learning

[5] J. Clark, “Import AI 431: Technological Optimism and Appropriate Fear,” Oct. 13, 2025. [Online]. Available: https://jack-clark.net/2025/10/13/import-ai-431-technological-optimism-and-appropriate-fear/

[6] M. Tegmark, Life 3.0: Being Human in the Age of Artificial Intelligence. New York: Knopf, 2017. [Online]. Available: https://www.amazon.com/Life-3-0-Being-Artificial-Intelligence/dp/1101946598

[7] Species Documenting AGI, “MIT Explains the 12 Possible Endings for AI,” YouTube, Mar. 2026. [Online]. Available: https://www.youtube.com/watch?v=FLcrvMfHUJM

[8] D. Patel, “Richard Sutton: Father of RL thinks LLMs are a dead end,” Dwarkesh Podcast, YouTube, 2025 (succession discussion at 54:52). [Online]. Available: https://youtu.be/21EYKqUsPfg?t=3292

[9] Wikipedia, “Vasily Arkhipov.” [Online]. Available: https://en.wikipedia.org/wiki/Vasily_Arkhipov

[10] Encyclopaedia Britannica, “Stanislav Petrov.” [Online]. Available: https://www.britannica.com/biography/Stanislav-Petrov