I’m one of these people. Simple answer? It’s the documentation. I’ve given k8s an honest try, but it honestly feels like the “draw the rest of the fucking owl” meme, starts way basic then gets wayy too hard without explaining anything in between.

Meanwhile docker is 1 file, 1 command to get started.

Edit: I just realized you were talking about k3s, not k8s. Is that something different somehow? Google says it’s a “k8s” distribution? WTF would that be?

Even then, Claude can also be a woman’s name in some cultures.

“Do not hallucinate”, lol… The best way to get a model to not hallucinate is to include the factual data in the prompt. But for that, you have to know the data in question…