News for hackers. Stuff that matters.

A comment on Large Language Models Are More Persuasive Than Incentivized Human Persuaders. Parent comment

Re:Large Language Models Are More Persuasive Than Incentiviz... by AlexCoventry on Saturday May 17, 2025 @10:37PM (#44018571)
It depends on the AI. ChatGPT's higher models (o1-pro/o3/o4-mini-high) have some kind of limited capability to detect errors in the user's thinking, and have relatively few hallucinations.
[ Parent | on HN ]
Re:Large Language Models Are More Persuasive Than Incentiviz... by UltraSane on Sunday May 18, 2025 @02:33AM (#44019435)
I've had fun debates about things like p-zombies with Gemini 2.5 Pro
[ Parent | on HN ]
Re:Large Language Models Are More Persuasive Than Incentiviz... by energy123 on Sunday May 18, 2025 @12:49PM (#44022632)
o3 have twice the hallucinations of o1 according to their own hallucination benchmark
[ Parent | on HN ]