I wrote this for another forum (9fans) as a few individuals are overly worried about AI use. I thought I'd share it here as well.
— —
I understand the frustration, especially if what you saw at IWP9 felt less like engineering than the outsourcing of thought. There is a difference between using a tool to assist reasoning and surrendering reasoning to the tool itself. Many people blur that distinction, and the result damages both the quality of the work and the culture surrounding it.
Your concern about human scale systems is legitimate. One reason Plan 9 and 9front continue to attract serious people is their clarity. The systems remain small enough for an individual to hold substantial portions of them in mind at once. That matters. When complexity exceeds human comprehension, engineers risk becoming mere operators of machinery they no longer fully understand.
I also agree that many vendor incentives surrounding AI are unhealthy. Much of the marketing centers on labor substitution, accelerated production, and the reduction of human involvement. It is understandable that people who value craftsmanship, judgment, and maintainability react against that. Some advocates speak as though human understanding itself is an obstacle to efficiency. That framing poisons the discussion from the outset.
At the same time, it is important to separate misuse from the tool itself. A person may misuse a compiler, debugger, framework, code generator, or search engine as readily as an LLM. The real failure occurs when the operator abandons independent judgment and ceases to take responsibility for understanding the result.
That distinction matters because otherwise the discussion risks becoming a purity test in which appearances outweigh competence. A person may write every line manually and still produce systems they do not truly understand. Another may use automation selectively while remaining fully capable of explaining, defending, maintaining, and correcting the work. The proper standard is understanding, authorship, and accountability.
Disclosure for conference presentations seems reasonable. Expecting presenters to understand and defend their work is reasonable. Rejecting low effort slop is reasonable. But I would hesitate to make tool usage itself the central criterion, because doing so may encourage concealment rather than responsibility.
Good engineering has always required thought, restraint, curiosity, and accountability. No tool alters that requirement. If a person chooses to use such a tool, he should use it carefully and remain capable of standing behind the work as his own.
