Is that a subjective feeling or is this statement backed up by something? The most recent trustworthy research on the topic disagrees: metr.org/blog/2025-07...

great, it automates the feedback loop to either get it right on the second try or hit a roadblock. It's a bold assertion of you to posit that it will definitely surely be able to get it right if it only gets access to the error messages without human intervention.

fell right into the obvious trap

Jaaa, das online Angebot von denen ist natürlich auch praktischer, gerade bei Sachen, bei denen es häufiger Tests gibt, weil man dann alles zu einer Produktkategorie auf ein mal sieht, aber die Magazine sind auch nicht schlecht

they're not harmful, sure. but if you're making the case for productivity, made-up stuff you'll have to fix yourself isn't exactly a productivity-booster.

unsere Stadtbibliothek hat die Magazine auch zur digitalen Ausleihe (kann halt auch nur x mal gleichzeitig ausgeliehen werden, aber abgesehen von der neusten sind die meisten immer verfügbar), vielleicht gibt's das bei euch ja auch?

I wouldn't trust genAI with more complex tasks than simple python scripts to automate a chore but I can also write those myself and those 15 saved minutes really aren't worth the implications

even if I base this purely on results and ignore everything from ethics to environment, I do not see the pros in a thing that hallucinates APIs and libraries that don't exist and therefore produces non-working code (and if it works it's likely to be a maintenance nightmare)

🤯 that tracks!!!

Goon Cave! At the Earth's Core is my favorite band

just looking at inaba raw hell yeah