882

Censorship is getting out of hand (lemmy.zip)

submitted 7 months ago by Maven@lemmy.zip to c/programmer_humor@programming.dev

87 comments fedilink hide all child comments

you are viewing a single comment's thread
view the rest of the comments

[-] ilinamorato@lemmy.world 151 points 7 months ago

We will never solve the Scunthorpe Problem.

[-] GeorgimusPrime@lemmy.world 44 points 7 months ago

It's a clbuttic

[-] Hexarei@programming.dev 7 points 7 months ago

Truly in a clbottom of its own

[-] SatouKazuma@programming.dev 25 points 7 months ago

Hasn't it been proven unsolvable?

[-] Valmond@lemmy.world 53 points 7 months ago

Impossible. There is always some mf named like cum-sock, smh

[-] ImplyingImplications@lemmy.ca 33 points 7 months ago

some mf named like cum-sock

Excuse me? My family BUILT this country!

[-] prowling4973@programming.dev 8 points 7 months ago* (last edited 7 months ago)

Or Grab-her.

[-] ilinamorato@lemmy.world 9 points 7 months ago

Proven? I don't think so. I don't think there's a way to devise a formal proof around it. But there's a lot of evidence that, even if it's technically solvable, we're nowhere close.

[-] elvith@feddit.org 14 points 7 months ago

Have you tried adding a few more kilobytes of regex?

[-] theterrasque@infosec.pub 6 points 7 months ago

Or a few more gb of LLM?

[-] elvith@feddit.org 4 points 7 months ago

I swear, I just need 4-5 more graphics cards to solve this!

[-] CanadaPlus@futurology.today 11 points 7 months ago

I mean, you could just use a vaguely smarter filter. A tiny "L"LM might have different problems, but not this one.

[-] tja@sh.itjust.works 14 points 7 months ago

So a TLM?

[-] CanadaPlus@futurology.today 6 points 7 months ago

TJA suggests a TLM.

[-] tja@sh.itjust.works 6 points 7 months ago

¯⁠\⁠_⁠(⁠ツ⁠)⁠_⁠/⁠¯

[-] Carighan@lemmy.world 6 points 7 months ago

Awww, it's trying its best!

[-] ilinamorato@lemmy.world 2 points 7 months ago* (last edited 7 months ago)

Indeed; it definitely would show some promise. At that point, you'd run into the problem of needing to continually update its weighting and models to account for evolving language, but that's probably not a completely unsolvable problem.

So maybe "never" is an exaggeration. As currently expressed, though, I think I can probably stand by my assertion.