Models like GPT-3 and early LLMs still hallucinate frequently, which caused early testers to dismiss them after a single bad output, creating selection bias where only people willing to tolerate failure continued using them.
factualpending
Speaker
Winston StrawnEvidence Quote
“Super risky because you know a decent amount of time it would hallucinate and get something wrong. And if it hallucinated the lawyer was like okay this is useless like get me off this demo. But when it worked it worked super well.”
Created: 8/10/2026, 11:28:11 PM
Your notes
Loading notes...