News for hackers. Stuff that matters.

A comment on The Waluigi Effect: an explanation of bizarre semiotic effects in LLMs. Parent comment

Re:The Waluigi Effect: an explanation of bizarre semiotic ef... by donnowhy on Monday March 06, 2023 @05:44PM (#35048849)
sounds to me like a wave with a positive and a negative part.

which IMO is what drives constructive/destructive interference in waves.

my take away is that any LLM that can behave "good" must also be able to behave "badly"; philosophically, because it's not possible to encode "good" without somehow "accidentally" but unavoidably also encoding "bad/evil".

This is well aligned with the rest of my understanding about the nature of reality including it's mathematically determined limitations (diagonals, infinities, paradoxes) and so on.

[ Parent | on HN ]
Re:The Waluigi Effect: an explanation of bizarre semiotic ef... by cs702 on Tuesday March 07, 2023 @06:35AM (#35054194)
> my take away is that any LLM that can behave "good" must also be able to behave "badly"; philosophically, because it's not possible to encode "good" without somehow "accidentally" but unavoidably also encoding "bad/evil".

That's a really good non-technical summary of the OP's hypothesis. Thanks!

[ Parent | on HN ]