Narrow finetuning can produce broadly misaligned LLMs (emergent-misalignment.com) 10 points by foweltschmerz 1y ago ↗ HN
[–] sylware 1y ago ↗ Is anybody pointing on the fact that "alignment" is brain-washing? [–] achierius 1y ago ↗ Is teaching a child? Is talking with your friend? Is punishing a criminal?Or, more pointedly, what about training the model in the first place? Why do you pretend that AI are somehow "people" with a "natural tendency" we're overriding?
[–] achierius 1y ago ↗ Is teaching a child? Is talking with your friend? Is punishing a criminal?Or, more pointedly, what about training the model in the first place? Why do you pretend that AI are somehow "people" with a "natural tendency" we're overriding?
3 comments
[ 0.85 ms ] story [ 17.1 ms ] threadOr, more pointedly, what about training the model in the first place? Why do you pretend that AI are somehow "people" with a "natural tendency" we're overriding?