jaschop

joined 2 years ago
[–] jaschop@awful.systems 24 points 3 days ago (5 children)

Bro, what's this (1+x)(1+x²)(1+x⁴)... ellipsis bullshit? You gotta get in on that big Π action!

[–] jaschop@awful.systems 6 points 1 week ago (1 children)

--yes-i-really-know-what-im-doing would also serve. This juvenile bullshit one should grow out of in their 20s at the latest.

[–] jaschop@awful.systems 17 points 2 weeks ago* (last edited 2 weeks ago)

I've never seen that, but it was my first guess as well. People here are already doing Doner-boxes with fries, meat and sauce in a takeout box, I'm assuming for people under doctor's orders to eat a veggie-free high-grease and high-sodium diet.

[–] jaschop@awful.systems 7 points 2 weeks ago

Derping around on TF2 trade servers. Not even trading. Just not being in the mood for a competitive game and rather fucking around with randos.

[–] jaschop@awful.systems 2 points 2 weeks ago* (last edited 2 weeks ago) (1 children)

Sorry for being late to the party, but I had to give my two cents on AI detectors. (specifically GPTZero & co., watermarking is different) I think it's a bad idea to recommend them. These companies are overpromising their accuracy and essentially selling false certainty. This will also not fade over time since the inaccuracy is baked into the tech.

Currently the performance may be good for basic use cases, but recent papers seem to indicate you can still dazzle them by burning more tokens or explicitly prompting against detectors. But from what I understand about GenAI training* this is an eternal cat & mouse game, since the AI vendors don't want their output to be detectable.

what I understand about GenAI trainingA basic technique of training a generative model is training a discriminative model alongside it. The gen model tries to synthesize output and the discriminative model tries to distinguish it from actual training data. The models are then trained alongside each other, building better and better discriminator models to sus out better and better generative ones.

If these detector companies had some kind of breakthrough (better techniques, better datasets, whatever) to sus out generated output, they would be bought up to improve GenAI training. More likely, they don't. Maybe they can dial into some typical usecases for specific models and perform highly there, but that will fail to generalize for the edge cases.

[–] jaschop@awful.systems 33 points 3 weeks ago (4 children)

Debian dads of lemmy, join me in an approving grunt.

318
ads rule (awful.systems)
submitted 1 month ago* (last edited 1 month ago) by jaschop@awful.systems to c/onehundredninetysix@lemmy.blahaj.zone
 

moments that make you think, things sure could suck even harder if I wasn't such a nerd

[–] jaschop@awful.systems 31 points 1 month ago (1 children)

Isn't evolution beatiful? Each of these specimen has adapted to fit a highly specific niche, it's knob optimized to predate a specific style of trouser pocket or even certain belt loops.

[–] jaschop@awful.systems 4 points 1 month ago* (last edited 1 month ago) (1 children)

This is bullshit and not how pro-Israel influence in US politics works. Kirk was broadly accepted by the republican mainstream and didn't stray too far from the party line. In fact he was for a while harrassed by Groypers in part because he refused to adopt bald-faced anti-semitism into his rethoric. This theory might well have originated in those circles.

[–] jaschop@awful.systems 25 points 1 month ago (1 children)

Yes we are. Bitches (a.k.a. my wife) be thirsting for this dad bod.

[–] jaschop@awful.systems 1 points 1 month ago

Wow, no upvotes. None of you did, then? Shame, it really holds up well. And you will get my previous comment.

(DM me if you need access.)

[–] jaschop@awful.systems 2 points 1 month ago* (last edited 1 month ago) (2 children)

So, who else did rewatch Master and Commander: The Far Side of the World recently?

[–] jaschop@awful.systems 6 points 1 month ago* (last edited 1 month ago) (4 children)

This rational animation channel seems to focus pretty non-stop on AI-risk crank talking points. Can't tell by skimming if they're engaging critically, but it seems unlikely.

And now you're apparently inspired to make a cutsiefied rendering of a slop machine running prompts, which is likened to a child being encouraged to explore.

I don't wanna shit on the art or anything, but it seems like the general AI psyop is getting to you a little.

 

Leftist streamer Vaush applies his unifying "everything I don't understand is Epstein shit" thesis to the fediverse.

Article from the clip: https://medevel.com/why-the-fediverse-isnt-mainstream-and-why-it-might-be-the-future-of-safe-private-social-media/

 
 
view more: next ›