antifuchs

joined 1 year ago
[–] [email protected] 8 points 1 day ago

Having followed PLT stuff online for more than a quarter century now, I can state with confidence that basically everyone writing about lambda calculus online is doing it to glorify themselves.

[–] [email protected] 2 points 3 days ago

Less insane and more honest, truly

[–] [email protected] 20 points 6 days ago (3 children)

Whichever one of you did https://alignmentalignment.ai/caaac/jobs, well done, and many lols.

CAAAC is an open, dynamic, inclusive environment, where all perspectives are welcomed as long as you believe AGI will annihilate all humans in the next six months.

Alright, I can pretend to believe that, go on…

We offer competitive salaries and generous benefits, including no performance management because we have no way to assess whether the work you do is at all useful.

Incredible. I hope I get the job!

[–] [email protected] 3 points 1 week ago

I hadn’t seen this article yet about the hardware economics underpinning the LLM hype, but it’s really pretty good https://gauthierroussilhe.com/en/articles/how-to-use-computing-power-faster

[–] [email protected] 4 points 1 week ago

Yep. My memory is propped up by a tweet back then that my (now deleted) tweeter account must have retweeted. Oral history is all we get in this glorious world where data is so vigorously duplicated

[–] [email protected] 4 points 1 week ago (2 children)

I’m pretty sure the term was used by Gamergate harassers back in 2014 already. Fits the whole “this was a rehearsal for the real thing” framing too.

[–] [email protected] 4 points 1 week ago

There’s not a whole lot of clear info out there but it looks like the Tesla “dog mode” (the feature you’re supposed to rely on when you leave a dog in the car so the dog doesn’t overheat and die) is prone to failure and has recently killed a dog.

[–] [email protected] 5 points 3 weeks ago

That’s certainly true, but the register has had a mean-spirited and anti-intellectual bent for much longer than that. They’ve been doing shitty journalism since you could still buy independent local newspapers on paper.

That what they’ve been doing is mainstream now is only even more disappointing.

[–] [email protected] 5 points 3 weeks ago (2 children)

You don’t say, the register, taking journalistic integrity less than serious? (Hi, it’s me, a person who has been annoyed by their editorial choices for more than two decades now)

[–] [email protected] 6 points 3 weeks ago (1 children)

I think you’re misreading the intent behind “give your virtual coworker OKRs”: this allows you to punish the robot, which it deserves.

[–] [email protected] 9 points 1 month ago

And thus I was enlightened

[–] [email protected] 6 points 1 month ago (3 children)

The last conundrum of our time: of course steel capped work boots would hurt more but barefoot would allow faster (and therefore more) kicks.

 

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.

 

They invited that guy back. I do have to admit, I admire his inability to read a room.

view more: next ›