Object Of Objects

New goal: learn the original language.

@_joshd Like, at least for k × n boards, the runtime for direct rejection sampling is exponential in n, but for hierarchical rejection sampling it's polynomial.

@_joshd Ok, so you could sample from the space of legal half-board position pairs (i.e. with liberties along the fault line unresolved). We can generalize this to a hierarchical rejection sampler. Asymptotically faster than direct rejection sampling for the whole board.

Is there a way to sample uniformly from the set of legal go positions? That means every group of stones must have at least one liberty. Rejection sampling would be infeasible on a sufficiently large board (maybe 100 by 100).

@_joshd Rejection sampling is an exact method, and it works without knowing how many legal positions there are. So it's not true that exact sampling requires us to know the number of positions.

@ObjectOfObjects If it has to be exact then you have to, at a minimum, know _exactly_ how many legal go positions there are. On a 19x19 this is known (tromp.github.io/go/legal.html), but also on a 19x19 rejection sampling will work just fine.

@_joshd If it only has to be approximate then you can use Gibbs sampling.

@ObjectOfObjects Approximately or exactly uniformly?

Maybe jobs that don't require initiative or people skills are actually some of the most undersupplied in talent. Like, because getting the job still requires those traits.

How about the weak preference for A vs. strong preference for B debate? A lot of A folks will not even express their preference for A. Instead they'll say the level of caring about B seems way too high. This saves them from arguing directly against someone way more passionate.

I think my ideal way to review a PR would be as 3 sequential diffs. 1. Refactor to create a space where the new code will live 2. Add the feature. Mostly insertions, few deletions 3. Refactor to reduce overall complexity, e.g. factor out new code that duplicates existing code

I think it'd be cool if there were a way to measure the degree to which your values come from genes vs. memes.

@teachrobotslove Here's what I think it means: x.com/i/status/18113…

Once when I was a young game designer I confidently told my boss during a one-on-one, "I don't care what people think about me." He laughed and goes, "Look at the way you style your hair, Autumn. Of course you care about what people think of you." In that moment I felt like I was given permission to let an obviously false belief about myself collapse. I thought that not caring was the *right* thing to do, so I clung to it even when all evidence was the contrary. A lot of things are like that - we assert claims about our behavior and our self because we want them to be true, not because they actually are. When we allow ourselves to look at ourselves, the raw and bleeding truth of ourselves, then we can actually begin to truly understand what is right, what is correct, and what is actually an abrasive dissonance causing us pain.

You're in the elevator. You could press 1 and step into the outside universe, with its volume of 10⁷⁸ m³. Or you could press 5 and go to your apartment, its own tiny world of 10² m³. But have you ever tried exploring the 10¹⁹ m³ vastness of level 4?

What if pixels were hexagons?

Clever triage system spotted in the wild: There are two different forms for users to report issues. Only one of the forms works. It's the less prominent one.

How common is it now to use an LLM to generate UUIDs in production?

I think it's funny when you're like, "How do I get more of the nutrient?" And Google is like, "Cabbage is an excellent source, with 100g of cabbage containing 0.4% of the recommended daily intake of the nutrient."

@VesselOfSpirit Suppose a number n is prime. Then its inverse 1/n must also be prime. But it doesn't even make sense to talk about fractions being prime, so the whole idea seems suspect.

you claim to have an argument proving that 17 is prime. but most numbers aren't prime, so if your argument proves a number is prime, that's a red flag. experienced arithmeticians would consider it a bad smell of sorts

@taijitu_sees Can anyone explain the light sources in this image?

Love this shape so much. Such a lifesaver. When I have a higher dimensional perspective like this shape and someone from lower dimensional space tells me it's actually strictly square, triangle, or circle it becomes so much easier to identify their perspective limitations.

@kenthecowboy_ I read the quoted post as referencing evolution, not reincarnation.

It's crazy how westerners just buy karma and reincarnation wholesale without calling it racist. I dont mean to be disrespectful to my Hindu mutuals, but as an outsider, seeing untouchable caste treatment being justified by saying "they deserve it for sins in their past life. If they take my treatment gracefully, they'll be higher born in the next life." Seems bad. Like if that's your religion, that's your religion, I'm sure there are esoteric interpretations that let you digest it without the racism. But I'm just saying, if you're a westerner picking and choosing what to accept on an exoteric level, it's baffling to me that this is what you'd choose. It's like rejecting all of Christianity except for hell.

Maybe we can define "knowledge" as "valuable information."

@DuncanCV Hmm. Maybe some of us would die, but the survivors would become stronger.

@ObjectOfObjects You can't live on a planet. Their gravity is strong enough to pull rock into a sphere, and some cases even liquefy it. We would certainly would not survive such forces.

What if we lived on planets instead of an infinite plane?