Post #2649402 · Elektrine

2026-02-08 22:29 UTC

@varx@infosec.exchange If your question is "Can LLMs generate a production-ready browser that I would trust today?" then the answer is obviously no. However, the fact that it can get within spitting distance with a single prompt should give one pause. It's easy to see how this story ends. More important I think is your first point – "good enough" or "worse is better" has been a defining trait of most software for a long time. I think we can both agree that we'll see a lot of crap software shipped this year. 😆

Replies (1)

@varx@infosec.exchange 2026-02-12 23:20

@nolan@toot.cafe I'm not much into betting, but I wonder if you'd like to make a prediction. What do you think is the probability that six months from now, you would be able to prompt an LLM to create a program with the following constraints? - Complexity on the order of IndexedDB - Not reimplementing something that already exists (I have several reasons for this constraint) - Accordingly, not driven off of a pre-existing test suite - High enough quality that you would risk your family's privacy and security on it - Maintainable code (either by human or further LLMs) - Takes at most a workday for prompting + manual fixups - API costs of at most $500 Or, how far out do you think it would be before this seems 90% likely to work? (Feel free to quibble with any of my constraints. I chose those because that's what it would take to impress me.)

Open ##2649403