antifuchs

joined 2 years ago
[–] antifuchs@awful.systems 4 points 11 hours ago

Aphyr weighing in with an ai position post:

Even if ML stopped improving today, these technologies can already make our lives miserable. Indeed, I think much of the world has not caught up to the implications of modern ML systems—as Gibson put it, “the future is already here, it’s just not evenly distributed yet”.

[–] antifuchs@awful.systems 7 points 6 days ago

For my own services I’m aiming for .999999% of uptime

[–] antifuchs@awful.systems 13 points 6 days ago

You just gotta appreciate that the most deterministic part of the code base is the predatory lootbox system.

[–] antifuchs@awful.systems 8 points 6 days ago (8 children)

Hold on now, the uptime number contains two digits that are nines! The image itself has four nines in total!

[–] antifuchs@awful.systems 6 points 6 days ago (5 children)

This thread by Johnny reading (skimming on a phone, hah) through it is really good.

If only literally any human with context and a small screen to look at the bigger picture was involved with decisions around taking this to production, it would … still be bad but only on a societal level.

[–] antifuchs@awful.systems 5 points 6 days ago

Heh. Who goes AI?

Never love the need for these parlor games, but it’s a good one.

[–] antifuchs@awful.systems 7 points 1 week ago

Make no mistake: Worm is not a webcomic. It is words. So many fucking words. (I made it about 30% in before realizing I could do literally anything else with my time).

[–] antifuchs@awful.systems 5 points 1 week ago (3 children)

Ah, of course! I come from a corp culture where co-workers were also called customers (“users” actually but you know)

[–] antifuchs@awful.systems 7 points 1 week ago (5 children)

Yeah, almost the entirety of use cases for genai chatbots is contempt for one’s customers.

[–] antifuchs@awful.systems 4 points 2 weeks ago

I’m like 60% sure he will self-incinerate when exposed to sunlight

[–] antifuchs@awful.systems 6 points 3 weeks ago

My heart goes out to my fleeting online acquaintance who’s seemingly but reliably two years too late to hype a trend. 2024 it was blockchain/cryptocurrencies that he tried pushing, now he’s saying AI technology companies are here to stay.

Somebody’s gotta buy the reverse reverse Cramer index I guess.

[–] antifuchs@awful.systems 3 points 4 weeks ago

Yeah, soatok has been doing really good work there. It’s disappointing that the matrix folks don’t seem to take it seriously.

 

Got the pointer to this from Allison Parrish who says it better than I could:

it's a very compelling paper, with a super clever methodology, and (i'm paraphrasing/extrapolating) shows that "alignment" strategies like RLHF only work to ensure that it never seems like a white person is saying something overtly racist, rather than addressing the actual prejudice baked into the model.

 

School student tells AI to put 20 other students’ faces on nude pictures, shares them in chat; it takes months for anyone including the school administrators to act because of some extremely, uh, dubious loophole.

If someone does that in photoshop, it’s a crime; if they do it in AI pretending to be photoshop, it’s somehow not. Gotta love this legal system’s focus on minor technicalities rather than the harm done.

 

They have Nik Suresh (the author) on, as well as Robert Evans. I haven’t listened to it all yet, but it’s fun so far.

view more: next ›