The Picard Maneuver@lemmy.worldM to Lemmy Shitpost@lemmy.world · 1 month agoDoesn't look like anything to me.lemmy.worldimagemessage-square33linkfedilinkarrow-up15arrow-down10
arrow-up15arrow-down1imageDoesn't look like anything to me.lemmy.worldThe Picard Maneuver@lemmy.worldM to Lemmy Shitpost@lemmy.world · 1 month agomessage-square33linkfedilink
minus-squareDesolateMood@lemm.eelinkfedilinkarrow-up1·1 month agoIt’s been trained to generate images that it thinks* can’t be distinguished from reality
minus-squareLarmyOfLone@lemm.eelinkfedilinkarrow-up0·1 month agoAnd if it could distinguish better, it could also generate better.
minus-squareNatanael@infosec.publinkfedilinkarrow-up0·1 month agoNot necessarily, but errors would be less obvious or weirder since it would spend more time in training
minus-squareLarmyOfLone@lemm.eelinkfedilinkarrow-up0·1 month agoWeirder? Interesting, like how for example?
minus-squareNatanael@infosec.publinkfedilinkarrow-up1·1 month agoWeirder in that it gets better at “photorealism” (textures, etc) but subjects might be nonsensical. Only teaching it how to avoid automated detection will not teach it to understand what scenes mean.
It’s been trained to generate images that it thinks* can’t be distinguished from reality
And if it could distinguish better, it could also generate better.
Not necessarily, but errors would be less obvious or weirder since it would spend more time in training
Weirder? Interesting, like how for example?
Weirder in that it gets better at “photorealism” (textures, etc) but subjects might be nonsensical. Only teaching it how to avoid automated detection will not teach it to understand what scenes mean.