True that!

[bbatsov@hachyderm.io]

RE: https://mastodon.social/@GeePawHill/116105595552650624

True that! I've seen just as much bad written by humans, as by LLMs…

[riffraff@mastodon.social]

@bbatsov imo there's a difference: I've seen (and written) a ton of bad code but that "works".
Like, 1000 line functions with global state which gives the right results.
OTOH, LLMs still produce, relatively often, code that does not work and then pretend it does.
Which sometimes people do too, but those are the people you want to fire.