3 ms·
Haha, I suppose the problem is that LLM outputs are unreliable yet presented as authoritative (disclaimers do little to counteract the boffo confidence with whi
by rectang 11mo ago
Haha, I suppose the problem is that LLM outputs are unreliable yet presented as authoritative (disclaimers do little to counteract the boffo confidence with which LLMs bullshit) — not that they are unreliable in unpredictable ways.
- johnisgood 11mo agoPresented as authoritative by its users. I mean there are very obvious disclaimers and people just ignore it.