cm0002@lemmy.world to Technology@lemmy.worldEnglish · 1 day agoAI models routinely lie when honesty conflicts with their goalswww.theregister.comexternal-linkmessage-square106linkfedilinkarrow-up1548arrow-down124
arrow-up1524arrow-down1external-linkAI models routinely lie when honesty conflicts with their goalswww.theregister.comcm0002@lemmy.world to Technology@lemmy.worldEnglish · 1 day agomessage-square106linkfedilink
minus-squareWanderingThoughts@europe.publinkfedilinkEnglisharrow-up1·1 day agoA lot of the improvement came from finding ways to make it bigger and more efficient. That is running into the inherent limits, so the real work with other models just started.
minus-squareNatanael@infosec.publinkfedilinkEnglisharrow-up2·1 day agoAnd from reinforcement learning (specifically, making it repeat tasks where the answer can be computer checked)
A lot of the improvement came from finding ways to make it bigger and more efficient. That is running into the inherent limits, so the real work with other models just started.
And from reinforcement learning (specifically, making it repeat tasks where the answer can be computer checked)