when i say something was possibly "caused by AI", i don't mean LLMs hallucinating or agents going rogue or whatever. i mean that it was caused by bad practices implemented thanks to AI psychosis in management. this can include laying off knowledgeable staff, pushing to prod without a thorough review, "build build build" etc. these are **human** problems.
3 Boosts
0 Likes
anyway, unless there is some massive course correction in corporate thinking, software is going to get worse and worse until people wish they could stop using it for important things.
over the last ~20yrs or so, software has basically worked, so corporate idiots have concluded it's a solved problem and they can get rid of a bunch of the people who were making it work behind the scenes, and now we're all going to pay for it. it's the same thing that happened with vaccines: you heroically solve the problem with **very** complex social and technological tools, and then the dumbest people in the world take over and conclude that the problem actually never existed.
fucking **Microsoft 365** was offline for 16 hours yesterday. if some hackers using AI did that, we would never hear the end of it. as it stands, it was probably bad engineering practices, same as this FAA outage, so everyone will just move on, while also gaslighting us that software and networks have always been bad. nope. with good engineers and proper engineering processes, this kind of thing has been **very** rare over the last 15yrs.