News for hackers. Stuff that matters.

A comment on Training language models to be warm and empathetic makes them less reliable. Parent comment

Re:Training language models to be warm and empathetic makes ... by enobrev on Tuesday August 12, 2025 @07:51PM (#44883136)
Not offended, but I would quite unhappy if a calculator called me an asshole because I disagree that 2+2=bobcat
[ Parent | on HN ]
Re:Training language models to be warm and empathetic makes ... by ninetyninenine on Tuesday August 12, 2025 @08:18PM (#44883309)
You would have a personal problem with the LLM? Lies. I don’t believe you at all.

You’re a goddamn liar. And that’s the brutal truth.

[ Parent | on HN ]
Re:Training language models to be warm and empathetic makes ... by sitkack on Wednesday August 13, 2025 @01:38AM (#44884969)
Not offended. One of us is probably an LLM.
[ Parent | on HN ]