thelaxiankey

I feel like the 2.49 to 2.6 update was a bit overblown, but the 2.8 update was pretty damn massive to me eyes

I mean besides the fact that it's, by construction, pseudorandom?

And beyond that the computation style is not turing-machine shaped, even remotely (although it's surely turing complete, but that doesn't mean the flavor is that of a turing machine)

Speculation on construction and development is absolutely fine and maybe even good! I don't have a problem with a bit of rent-seeking if you went ahead and built something.

However... speculation on the underlying land, which is not fungible and one of the few things on this planet earth that is actually completely inelastic with a totally fixed supply, is bad. I won't say this is a popular tax, but it's not unheard of.

I think a really great example here is Singapore, which effectively doesn't have private land ownership and has been pretty spectacularly well developed in the last few decades.

Yes, there is a complexity limit, but it is fuzzy and not perfectly well defined (and that's ok). Modern LLMs ride that line in a wild and interesting way.

It's really fun having become a biologist in recent years, and knowing that there's an entire field that uses living organisms (yeast or neutered E coli) as protein factories. I am bringing it up a bunch here because not only are these things used as tools, but they're literally alive!

It's a tool but it's not a strict computer program, and the language we use for computer programs doesn't capture their behavior well. In squishier fields like bio, where model organisms are used as living protein factories, there's a reason we don't call maladaptive behavior a bug. When dealing with useful but complicated black-box systems, we should feel free to reach for any metaphor we please.

Would it make you feel better if the user said 'It'll make significant changes to how the code works. In the COT, it brings up explanations that resemble anxiety about copying code, and my suspicion is that this is the proximal cause of the changes it's making.' -- this is inarguably true, and very likely what the guy you're responding to meant.

Stockholm is colder, further north, (though admittedly less hilly).

I can't argue that the bike infra improvements in Seattle haven't made a huge difference, but I'm not at all convinced this is because it is intrinsically unlikeable.

For what it's worth, the environmentally incredibly bikeable SD/LA has such bad infra that they have both fewer bikers and far more deaths.

This happens a lot in social sciences; more than once I've gotten my intuition "disproven" by a study only for that study itself to later be proven wrong.

However this study in particular is pretty bad, see the comment elsewhere about the screentime not spent not going down (possibly cause of extant classroom bans)

On the flip side, when measurements don't reveal something that is completely obvious, especially in social sciences, it is very possible the measurement is measuring the wrong thing.

In this case the prevailing theory is that the phones were already banned in classes, so the schoolwide ban didn't do anything to screentime. Which, duh.

Why not? Just cause it's not strictly in steady state doesn't mean we aren't approximating steady state

Personally, I don't have a problem with land ownership, i've got a problem with speculation and rent seeking. The point of a lease is you've gotta pay rent, which to my eyes is a good, easy to implement approximation of a land value tax.

Just fyi adhd is not considered a categorical disease ATM. This means that it is not necessarily a disease in the same sense that flu or schizophrenia are diseases. It's a disease of excess

sticking fluorescence (and luminescence) inside an organism of choice is not particularly hard! just add a GFP gene and you're done. if you can do any genetics at all, this is one of the first things to try.

People gave their lives so everyone could spend their life studying law?? The people you've got in mind gave their lives for a representative, not a direct democracy.

Laws are difficult and nuanced. Deciding between yes/no takes me a long time, because I am not a lawyer or a legislator. Even for them, again, legislating is a job for a reason... I know I like certain things, and dislike others, I know basically which direction I want things to go; however, deciding whether a concrete bill is a reasonable-enough implementation that has closed loopholes is way out of the scope of decisions I would trust myself, or most citizens, to make well. This, it feels like a waste.

For what it's worth I was a similar kind of child. My problems entirely went away with about 10 (or more) hours of exercise a week. This required a hobby I cared about and the right kind of friends, but since then things have been smooth

Not necessarily; these learning things are super non convex and it's not clear to me "where" they max out

Don't know about you, but personally I don't like direct democracy bc i have shit to do besides deciding which law to pass. My states' ballot is full of ridiculous questions every year and takes me days to do my research. I can't imagine doing this year round without getting paid.

it's funny to me how the success/not total implosion of htese companies is predicated on profitable, revolutionary-tier success, and that it's increasingly possible that the profits will never really materialize. pretty interesting move on China's part.

Yeah but it didn't do squat. The effective thing was a subtle prompt mod that boosted performance tremendously. I've been surprised by this forums unwillingness to read this thing...

I think your summary is not really correct, maybe I'm missing something. As far as I can read, the approximate takeaways are these:

* LLM chess play is hyper sensitive to the harness being used (see: the section on regurgitation), and only mildly sensitive to fine tuning

* The best ELOs the author was observing were from 1500 to 1750 or so (circa 2024/25). Not grandmaster, but no longer incompetent monkey either.

Haha :/ one of the sad things for me is that prior to LLMs I had a fairly LLMish writing style. Kinda formal, lots of bullet points and em dashes, two instances before etc, etc. since the debut of gpts I've had to adjust my style to be marginally more opaque so that I don't get mistaken for an LLM

It's also how the San Diego trolley works (it's called a trolley, but it is almost a heavy rail with respect to how separated the tracks are everywhere except downtown). One of the cleanest, most effective rail systems I've seen in the US (though there are a lot of security in the wagons)

there are far more reams of chess moves than there are math papers. Lichess is pretty open...

But hey, they're actually good at chess if you prompt correctly so.... https://dynomight.net/more-chess/

I'm just dropping this all over this thread but you're unfortunately mistaken

https://dynomight.net/more-chess/

Sure: https://dynomight.net/more-chess/

Forgot to add the amazing follow up

https://dynomight.net/more-chess/

These "researchers" are less informed on LLM chess than random internet bloggers. The situation is much more interesting

https://dynomight.net/chess/

My suspicion is small phone users buy phones less often, and thus are less represented by the market. This may be borne out if a company randomly produces a small phone and it proves popular

are you an LLM? I know what gfp is lol

ah yeah, the intelligence claim was basically an afterthought in the actual paper, the smithsonian mag did a bad job with their reporting. the paper is just called: 'Evolution of a core ribosomal innovation in octopus', and I would argue the E coli thing is far from a non-sequitur in that context

the cross-organism increase in fidelity is pretty damn shocking. Ribosomes are one of the most optimized, slow-evolving things out there, and transplanting genes from animals to coli is already a very checkered endeavor.

this is literally not possible. while I'm excited for these kinds of things, alphafold is already imperfect, and this basically side-steps actual structural predictions. it sounds good, but frankly biology is not there yet. in practice not even alpha fold can tell me if my gfp fusion will work...