Fernando Borretti expected AI to change his professional work while leaving the pleasures of programming intact: learning languages, reading papers, writing essays, and making small open-source projects. His worry now is not simply that machines can do those things. It is that fewer people may need what another person contributes.
The garden needs a bazaar
- Intellectual work is not wholly private. It depends on a shared body of knowledge and people who use, challenge, and build on it.
- Recognition is more than a vanity metric. A citation, a useful reply, or a contribution to a repository tells someone their work helped another person. That exchange can sustain effort when intrinsic enthusiasm fades.
- More output need not mean more participation. An agent can grow your private garden of code without sending you into the bazaar to learn from or trade with other people.
- AI mediation can hide the contributor. A technical explanation might help a model answer someone’s question without its reader ever encountering the person who wrote it.
The distinctive argument is about the social conditions of motivation, not a productivity benchmark. Wanting your work to matter to other people is not necessarily an inferior motive that automation should purify away.
Where the conclusion outruns the evidence
Borretti moves from a recognizable loss of connection to much stronger premises: humans no longer write or read code, contributions become superfluous, and intellectual communities dissolve. The essay does not establish those as universal outcomes. Reduced demand for some kinds of work is not proof that every human audience disappears.
His warning is more useful as a design question: which interactions are we eliminating, and which can we deliberately preserve? Faster answers can replace an occasion for mentorship or friendship—even when the answer itself is good.
What the thread adds
The early Hacker News discussion puts specific relationships behind the essay’s abstract claim:
- sdevonoes — On asking colleagues for help instead of directing every question to an agent: “I don’t mind spending 2h helping a colleague debugging an issue. It’s fun, it creates bonds and we both benefit from it”. The lost interaction is part of onboarding and team formation, not just a slower search interface.
- alecst — On expertise as a reason friends used to call: “Without the question, they might have never called. But because they called, we usually ended up just talking about other things too. Life. Girls. Work.” They also describe continuing weekly meetings around a toy problem that an LLM could probably solve: the shared activity still serves a purpose beyond the result.
- chrisfosterelli — Describes peers outside the booster/doomer split: “those that recognize AI is super valuable for getting a lot of work done but also feel that suddenly the work is so much less exciting than it was before.” Usefulness and diminished enjoyment can coexist.
- tombert — Offers a deliberate boundary for personal projects: “I still write code by hand … just because I realized I was losing my intuition about software because I was outsourcing to Claude for everything.” Keeping a practice can matter independently of winning a speed comparison.
Where the thread pushes back on the premise
- mccoyb — Argues that language and compiler design could gain a different audience and purpose: “I think about those systems significantly more now, because I view them as ways to describe the search space for the agent”. That challenges redundancy, although it does not by itself solve the loss of human recognition.
- sanderjd — In reply to tombert, disputes reducing the new work to prompting: “The skill isn’t ‘prompting’, it is figuring out that to do, validating that it works the way you want, and thinking about how to best support future iterations of things you might want to do.” Their argument is that important engineering responsibilities persist across the tool change.
The unresolved question is whether AI-mediated work can preserve those occasions for human exchange deliberately, rather than treating every request for help as friction to remove. The thread supplies experiences and possibilities, not a measured answer about the future of intellectual communities.
This is an early slice of a thread reporting 14 comments when checked. Handles are pseudonymous; HN exposes no per-comment scores, and its API ordering reflects its own ranking rather than a public vote tally. These attributed views are not evidence of consensus.