Hacker News new | ask | show | jobs
by rootnod3 207 days ago
As if the language models currently would give a damn about copyright...
1 comments

The problem is they have to hide their sources due to copyright. So they train on copyright data but must obscure it in the output. Thus they must hide the sources of truth making it impossible to fact check them directly and the reason that hallucinations are so common and unavoidable in the current pattern.