clickbait in the forest

Lesswrong lurker STEMlord. Sentience comparable to that of LLMs. Socially moderately homosexual, fully homosexual otherwise. AFAB. 20s. Future Batman villain and/or crazy hermit, depending on the position and momentum of the stars. Hater at heart, lover at brain. I follow people who seem interesting

[mielivalta1]

A list of the non-self-explanatory tags on this blog and what they mean

a: art

ae: aesthetic

m: mine

r: to read later

s: science

I'll add links later. Will update as needed

[hexbugswallower]

I really enjoy the embryonic-stage xkcd comics from before randall munroe knew how to make them funny

[maudit]

[profeminist]

RIP Marjane Satrapi, author of the amazing graphic novels Persepolis about living during the fundamentalist revolution in Iran in the 70’s and 80’s. She also created the animated movie based on the graphic novels, which is where these gifs come from.

Gifset source

[missbliss12]

Reblogging in honor of Marjane Satrapi, one of THE great graphic novelists. Her comic Persepolis was a crucial text for shaping my belief that comics can deeply explore identity, culture, politics, and history.

[porkue-deactivated20230306]

At any given moment I'm thinking about this pic

[mielivalta1]

I love my grandpa. My grandpa loves Le Guin. So I'm reading Le Guin as a sort of ritual to get spiritually closer to him. Maybe I'm doing this being human thing right after all

[mielivalta1]

ANY FANDOM THAT I'M IN. NO MATTER HOW MANY TAGS I BLACKLIST. ALWAYS IN THE ALL TIME-TOP POSTS. I'M DDOSSING THIS SITE

[sirensxxsongs]

The best shot from the final holy shit

[soldierboys]

THE BOYS 5.08 Blood and Bone

[thefingerfuckingfemalefury]

This is how I want Absolute! Wonder Woman's "reunion" with the Amazons to go ;D

[argumate]

lol Claude

[eightyonekilograms]

Behind the Scenes Hardening Firefox with Claude Mythos Preview – Mozilla Hacks - the Web developer blog

New details about what we found, and how agentic harnesses are now able to reproduce real bugs and dismiss false positives.

Suddenly, the bugs are very good

Just a few months ago, AI-generated security bug reports to open source projects were mostly known for being unwanted slop. Dealing with reports that look plausibly correct but are wrong imposes an asymmetric cost on project maintainers: it’s cheap and easy to prompt an LLM to find a “problem” in code, but slow and expensive to respond to it.

It is difficult to overstate how much this dynamic changed for us over a few short months. This was due to a combination of two main factors. First, the models got a lot more capable. Second, we dramatically improved our techniques for harnessing these models — steering them, scaling them, and stacking them to generate large amounts of signal and filter out the noise.

I know this sounds like a scam to keep you constantly immersed in the hype machine, but it's just an unfortunate-but-true fact that if you're not following up-to-the-minute news on AI capabilities, your opinion about them is wrong.

[invertebrateinvert]

[raccoonmilf]

I don’t have a raccoon daughter because I’m a raccoon biologist. I have a raccoon daughter because I moved to South Africa because I had what I thought was a “prophetic dream” (I had scurvy and also an evil psychiatrist prescribing me the wrong meds) where I was in South Africa and there was billboard with a woman on it in a lab coat holding a red fox and a raccoon that said “Dr Foxy: This Could Be You!” And then in the dream I looked across the street and saw a billboard the said “Come Visit Hooters in South Africa.” And I woke up and was like “I know what I need to do” which was not “go to school to study native wildlife” but instead was “visit Hooters in South Africa.” But I didn’t want to go for just a little while because it was expensive and I didn’t like the idea of the long flight, but I knew I HAD to go to the Hooters in South Africa. So I figured it would be more economical to just go and finish art school there. Except COVID happened and I literally got trapped there and the hooters in the city I moved to had apparently been closed for years and also I got a concussion and when I went to the doctor they said I had scurvy. So I had to do intensive eating treatment where I drank a lot of fruit juice and also learned to eat macaroni that was shapes other than SpongeBob. And after listening to my yapping about raccoons, my therapist, who had never met a raccoon, told me I should get a raccoon to help me keep fresh fruits in the house. And so when I got back to America I found someone trying to get rid of one Facebook because it was apparently evil and bit her toddler and then I did.

And now I’m a raccoon biologist.

[derinthescarletpescatarian]

#thats exactly what id expect from anyone who's made it big post dashcon

I love how the phrasing of this politely obscures that raccoonmilf made Dashcon. Like a newcomer could come away thinking you mean "someone who made it big after Dashcon happened." But no.

[pnglove]

[hugsfromhugo]

A solar eclipse and the Milky  Way seen from the ISS

[shimmeringdroplet]

So far I've mostly not been super excited about AI math progress, but the recent solution to Erdős 1196 seems pretty crazy. I know approximately zero number theory, but an actual mathematician is saying things like "this AI proof is from The Book. Perhaps the first?" and "I care deeply about this problem, and I've been thinking about it for the past 7 years. I'd frequently talk to Maynard about it in our meetings, and consulted over the years with several experts [...] not a question of low-visibility per-se. Rather, it seems like a proof which becomes strikingly compact post-hoc, but the construction is quite special among many similar variations."

Even more surprising to me (though I again know nothing about the area) is the forum thread, where Terry Tao is talking about how the proof might shed light on the "anatomy of the integers", whatever that means. This was the most memorable quote for me:

"In any case, I would indeed say that this is a situation in which the AI-generated paper inadvertently highlighted a tighter connection between two areas of mathematics (in this case, the anatomy of integers and the theory of Markov processes) than had previously been made explicit in the literature (though there were hints and precursors scattered therein which one can see in retrospect). That would be a meaningful contribution to the anatomy of integers that goes well beyond the solution of this particular Erdos problem."

I really don't know anything about this topic, but I'm curious if anyone here has thoughts on this; this seems like a pretty huge advance in AI math advancements.

[mielivalta1]

Poincare sausage machine......

Thus, be it understood, to demonstrate a theorem, it is neither necessary nor even advantageous to know what it means. The geometer might be replaced by the logic piano imagined by Stanley Jevons; or, if you choose, a machine might be imagined where the assumptions were put in at one end, while the theorems came out at the other, like the legendary Chicago machine where the pigs go in alive and come out transformed into hams and sausages. No more than these machines need the mathematician know what he does.

— Henri Poincaré

From 'Les Mathématiques et la Logique', Science et Méthode (1908, 1920)

[collapsedsquid]