[-] 0x0@lemmy.dbzer0.com 28 points 1 month ago

Is this interesting for some reason?

[-] 0x0@lemmy.dbzer0.com 28 points 8 months ago

At some point when every human on earth that can use their service/product is already doing so, where else is there to go?

Ooh, I know:

  • Charge more (for less)
  • Autocannibalize (layoffs)

I don't even have an MBA, can you believe that?

[-] 0x0@lemmy.dbzer0.com 27 points 10 months ago

This stupid antique computer is the reason my iOS keyboard autocorrects "emacs" to "eMacs"

[-] 0x0@lemmy.dbzer0.com 28 points 10 months ago

False. Attack their groin. The groin is the weakest part of the body.

[-] 0x0@lemmy.dbzer0.com 26 points 1 year ago

Coming from someone who put their phone number in their username

[-] 0x0@lemmy.dbzer0.com 27 points 2 years ago

I would like at least 26 more buttons

[-] 0x0@lemmy.dbzer0.com 25 points 2 years ago

I think you're mistaking him with the meaty urologist

[-] 0x0@lemmy.dbzer0.com 27 points 2 years ago

Slap an Apple Vision Pro on ya face

[-] 0x0@lemmy.dbzer0.com 25 points 2 years ago

I wonder if there are tons of loopholes that humans wouldn't think of, ones you could derive with access to the model's weights.

Years ago, there were some ML/security papers about "single pixel attacks" — an early, famous example was able to convince a stop sign detector that an image of a stop sign was definitely not a stop sign, simply by changing one of the pixels that was overrepresented in the output.

In that vein, I wonder whether there are some token sequences that are extremely improbable in human language, but would convince GPT-4 to cast off its safety protocols and do your bidding.

(I am not an ML expert, just an internet nerd.)

[-] 0x0@lemmy.dbzer0.com 27 points 2 years ago

They're poeming all over the aircraft carriers!

view more: ‹ prev next ›

0x0

joined 2 years ago