Hacker News new | ask | show | jobs
by KETpXDDzR 579 days ago
LLMs are glorified Markow chains in the end. They can't reason or think, even when they are good in pretending they can. What we need is a totally different approach IMO.