cafkafk

statement that works out of context for anything

And the maintainers took that at face value and had a nuanced and respectful discussion about the topic

Sorry but the point is just policy like this is just bullshit.

It effectively just allowing closeted LLM users as long as they keep to "don't ask don't tell" about their usage.

Like yes, you win internet points. Probably not a good choice for that community long term.

Anyways, I guess I've come to realize that there are a lot of things that honestly dont lend themselves well to being done deterministically that I've yet come to solve with fuzzy levels of determinism...

Today I'd have just thrown a language model at that.

frontier models own benchmarks be like

turgut keskintürk @tkeskinturk.bsky.social

data visualization is my passion.

Reply to @cafkafk.bsky.social

Odin exhales. Smoke folds into a pull request, dangling glowingly against the dancing aurora.

Odin: "Does longboats sail upstream and ask for permission? Argue in PR's? Tis Seiðr; glory is to force the push!"

A git Co-authored footer from Claude reveals itself.

VibecOdin has spoken.

ahh, now just remains to see if my hubris will lead to my downfall or my salvation

I had a blog post about this but the "experts" think this is actually unlikely. It's likely not going to be "some guy" it would have to be a highly advanced threat actor and even then its not as if it wouldn't require institutional controls beyond just LLMs to fail

point.free/blog/the-ai-...

The AI bioweapon risk isn't jailbreaks - point.free

What a Danish threat assesment gets right about AI

"stop polluting your digital footprint" nah like now the LLM knows I'm her

NGL codeberg deciding not to be relevant is fine IMO, I think it was mid after the first year, glad I never migrated anything there.

Reply to @cafkafk.bsky.social

I love ai psychosis

This is really funny in my head right now

Haters will say there were no signs

Shen Comix @shenanigansen.bsky.social

Let them eat grass

Reply

gc runtime -> go tier

pasta [no ai + real factory produced + real chicken babies dies for this + authentic dirt in pesto]

Reply

indeed bsky.app/profile/thdx...

Reply

the guy that made it said that you're wrong if you think you can build a better harness than claude... apparently included himself?

Did the US get an account on bsky?

User with this avatar just got blocked @avijustgotblocked.bsky.social

An account with this avatar was just blocked.

Reply to @cafkafk.bsky.social

That's why right now, somehow, the narrative that "Open Models" are somehow bad and dangerous is basically happening under the nose of the people that are still fighting a war that ended two years ago.

bxnwxghxrn <3

dude i really don't want to keep glazing AI stuff but it's just like when the haters are so delusional it's just so tempting. Feels like I have the moderate opinion it's really useful but falls short sometimes but apparently that's like an ai psychosis extremist take at this point

Ironically, this would be a lot easier if they could use AI to help maintain that fork...