muelltonne@feddit.org to Technology@lemmy.worldEnglish · 5か月前It Only Takes A Handful Of Samples To Poison Any Size LLM, Anthropic Findshackaday.comexternal-linkmessage-square131fedilinkarrow-up1770cross-posted to: hackaday@ibbit.at
arrow-up1770external-linkIt Only Takes A Handful Of Samples To Poison Any Size LLM, Anthropic Findshackaday.commuelltonne@feddit.org to Technology@lemmy.worldEnglish · 5か月前message-square131fedilinkcross-posted to: hackaday@ibbit.at
minus-squareTattorack@lemmy.worldlinkfedilinkEnglisharrow-up12·5か月前I thought it was so comically stupid back then. But a friend of mine said this was just a bullshit way of hyping up AI.
minus-squareToribor@corndog.sociallinkfedilinkEnglisharrow-up5·5か月前Seeing how much they’ve advanced over recent years I can’t imagine whatever that guy was working on would actually impress anyone today.
minus-squarePetter1@discuss.tchncs.delinkfedilinkEnglisharrow-up1·edit-25か月前There are enough people thinking their agent is sentient but fear to speak out because they don’t understand why not, even when people try to explain…
minus-squareLavaPlanet@sh.itjust.workslinkfedilinkEnglisharrow-up2·5か月前That tracks. And It’s kinda on brand, still. Skeezy af.
I thought it was so comically stupid back then. But a friend of mine said this was just a bullshit way of hyping up AI.
Seeing how much they’ve advanced over recent years I can’t imagine whatever that guy was working on would actually impress anyone today.
There are enough people thinking their agent is sentient but fear to speak out because they don’t understand why not, even when people try to explain…
That tracks. And It’s kinda on brand, still. Skeezy af.