36
Chat is this deep? (lemmy.ml)
submitted 1 year ago by to c/memes@lemmy.ml
9 comments fedilink hide all child comments
 

Reference: AI decides on trolley problem.

you are viewing a single comment's thread
view the rest of the comments
[–] 5 points 1 year ago (1 child)

What aspect in Deepseek's training may have created that result? Like would it be something in training or more likely the training data?

  • permalink
  • fedilink
  • source
  • hideshow 2 child comments
    •
  • [–] [S] 5 points 1 year ago

    Claude to me performed way better then the others on a moral level. I think this is due to training rather then the quality of data. Grok has more data then Claude and it’s an unhinged piece of shit.

  • permalink
  • fedilink
  • source
  • parent