Maven (famous)@lemmy.world to Programmer Humor@programming.dev · 3 months ago"prompt engineering"lemmy.worldimagemessage-square86fedilinkarrow-up11arrow-down10
arrow-up11arrow-down1image"prompt engineering"lemmy.worldMaven (famous)@lemmy.world to Programmer Humor@programming.dev · 3 months agomessage-square86fedilink
minus-squarexmunk@sh.itjust.workslinkfedilinkarrow-up1·3 months agoUsing another AI to detect if an AI is misbehaving just sounds like the halting problem but with more steps.
minus-squaremarcos@lemmy.worldlinkfedilinkarrow-up0·3 months agoLots of things in AI make no sense and really shouldn’t work… except that they do. Deep learning is one of those.
minus-squareMatch!!@pawb.sociallinkfedilinkEnglisharrow-up0·3 months agoGenerative adversarial networks are really effective actually!
minus-squareNatanael@slrpnk.netlinkfedilinkarrow-up1·3 months agoAs long as you can correctly model the target behavior in a sufficiently complete way, and capture all necessary context in the inputs!
Using another AI to detect if an AI is misbehaving just sounds like the halting problem but with more steps.
Lots of things in AI make no sense and really shouldn’t work… except that they do.
Deep learning is one of those.
Generative adversarial networks are really effective actually!
As long as you can correctly model the target behavior in a sufficiently complete way, and capture all necessary context in the inputs!