antifuchs

joined 11 months ago
[–] [email protected] 8 points 1 week ago

Which also isn’t about either of those things.

[–] [email protected] 12 points 1 week ago

My favorite meme message board on the internet keeps putting out bangers, but this one is extra good and relevant http://www.b3ta.com/board/11416629

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

Would you believe this prescient vibe coding manual came out in 2015! https://mowillemsworkshop.com/catalog/books/item/i-really-like-slop

[–] [email protected] 7 points 2 weeks ago

Also he wrote borderline anti-woke stuff back when doing that could still appear edgy and icky.

[–] [email protected] 12 points 3 weeks ago (1 children)

Really wonder if his own pictures are included in the training set (as negative examples, of course)

[–] [email protected] 7 points 1 month ago

So many projects and small websites I’m aware of are being overtaxed by shitty LLM scrapers these days, it feels like an intentional attack. I guess the idea of ai can’t fail, it can only be failed; and so its profiteers must sabotage anything that indicates it’s not beneficial/necessary.

[–] [email protected] 5 points 1 month ago

This “insight” tool gives a very incisive critique of the opinion-page journalism we’ve been seeing lately. Oddly enough, not the kind of critique you’d see printed on an opinion page.

[–] [email protected] 11 points 1 month ago

Just a couple more GPUs bro, I’m sure it’s around the corner

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

Ah, yes. The implication.

[–] [email protected] 8 points 2 months ago

Thanks. I simply switched to Fastmail over this bullshit. (“Simply” mileage may vary)

[–] [email protected] 15 points 2 months ago

They’ll react to it … by swerving further right.

[–] [email protected] 0 points 5 months ago (1 children)

Microsoft found a fitting way to punish AI for collaborating with SEO spammers in generating slop: make it use the GitHub code review tools. https://github.blog/changelog/2024-10-29-refine-and-validate-code-review-suggestions-with-copilot-workspace-public-preview/

 

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.

 

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 ›