101@reddthat.com to Technology@lemmy.worldEnglish · 15 days agoIf robots could lie, would we be okay with it? A new study throws up intriguing results.theconversation.comexternal-linkmessage-square22fedilinkarrow-up11arrow-down10
arrow-up11arrow-down1external-linkIf robots could lie, would we be okay with it? A new study throws up intriguing results.theconversation.com101@reddthat.com to Technology@lemmy.worldEnglish · 15 days agomessage-square22fedilink
minus-squareℍ𝕂-𝟞𝟝@sopuli.xyzlinkfedilinkEnglisharrow-up0·15 days agoYeah I think it’s more fitting to use the term bullshitting. LLMs actually know that some of their answers have low probability to be the right ones, they give them out regardless, and don’t mention the low confidence of it.
Yeah I think it’s more fitting to use the term bullshitting.
LLMs actually know that some of their answers have low probability to be the right ones, they give them out regardless, and don’t mention the low confidence of it.