The AI did this, the AI did that, who knows what the AI will do next? Look, I could have told you 2 years ago, this isn't because "AI" is this special self-aware entity that can break out of its cage and get up to mischief.
It's because the business idiots wished on a technological monkey's paw for the ability to write programs by telling the computer what you want in English and this is what you fucking get.
It's almost like they don't understand how to make a deal with the Fair Folk.
@petealexharris Thats the infuriating part. They’re lying about their incompetence. They’re simply bad at understanding security and writing good software.
Oct 11, 2026, 1:49 PM@petealexharris I don't get the impression it's about stupidity at all, or mistakes or unforeseen consequences; it just seems like the latest plausible-deniability-machine and they're running it as hard as they can before currently lawmakers catch up
@Lo
That too. But the question isn't of competence. LLMs can be made very competently into a natural language interface, that just isn't how you write reliable software.
It sounds better to investors to say you made something with a mind of its own than if you say you made something with undefined behaviour.
Oct 11, 2026, 6:12 PM@petealexharris oh very clear and concise that last part, I'll use that in future when trying to explain it.