scarmig · HackerLangs

há 4 horasdiscuss

LLMs have basic reasoning and a whole lot of memorization. Through that basic reasoning and pruned search, combined with piles of compute, you can prove lots of things. But the memorization of human failure prunes that possibility, and you need to expend effort convincing the LLM not to prematurely prune based on previous human failure.

scarmig

há 5 horasdiscuss

Two years from now, when a proof of the Collatz conjecture is announced and verified:

"No one cares about that! What practical applications does it have? It's pretty trivial anyways. More hype from the AI psychosis crowd."

scarmig

anteontemdiscuss

You can solve it with two heaps, which don't need to maintain a complete order. Or selection algorithms, as in the sibling comment (asymptotically better).

scarmig

há 4 diasdiscuss

Everything we do has side effects. Some have more than others; in the case of water, even heavy users of AI have a relatively small AI water footprint compared other things they do.

There are ideological perspectives--another comment in this chain declares that anything that an LLM does has zero value, purely by virtue of being from an LLM. There's not much to discuss there: maybe I find alfalfa useless and LLMs provide a lot of value, maybe you do the opposite. Short of violence, either immediate or delegated, the best way to resolve this conflict is to incorporate costs for resources like water appropriately (accounting for externalities), and let market participants bid to determine who gets access to it.

Datacenter owners are highly likely to find this a reasonable process, because they believe that what they are investing in and operating will provide a great deal of value on the market. And other industries (like your alfalfa) are far more likely to throw a fit at this process.

Or if you think alfalfa farms have fundamental, deep-seated water rights that don't extend to data centers, then data center operators can just purchase water from alfalfa farmers, and everyone involved will eagerly make that trade.

scarmig

há 5 diasdiscuss

The hysteria around water usage rests on people not knowing the scale of industrial civilization. First thing to do is compare any estimate of data center water usage with the water usage of almond farming. Or, if you want to focus on individual consumer choices, the water footprint of eating a hamburger.

scarmig

há 5 diasdiscuss

There's no statement one way or another about should in my comment; and, for what it's worth, my ideal would be an immediate global pause in AI research and development.

But the different terms imply different mental models of what LLMs are and can do. If you take two people, one who thinks of them as "artificial intelligence" and one as "stochastic parrots" (with all the implicit context and connotations of the individual words composing them), what mental model would have led to better predictions of LLMs' future circa 2020?

The "stochastic parrots" phrase is very dangerous in that frame. People read far more into what capabilities it implies are (im)possible than the narrow technical description the authors originally argued for. If all they are is spicy autocomplete or pastiche plagiarizers, there's nothing serious to worry about. And when an opposition gets stuck in a trough that mindlessly dismisses their future capabilities out of hand because of a bad mental model, it renders them ineffective at preventing the worst outcomes.

scarmig

há 5 diasdiscuss

Which frame inspires a more productive research program? Which has better predicted the trajectory of capabilities over the past five years?

scarmig

há 9 diasdiscuss

Back then, it sounded like making right-handed life was decades away. But with this work, couldn't you just as easily make this kind of synthetic cell right-handed?

scarmig

há 10 diasdiscuss

It's right up there with all the propaganda to convince the public that vaccines are good, or that drinking paint thinner is bad for you.

scarmig

há 10 diasdiscuss

It has to do with the scope of what they're discussing. It seems extraordinarily small: e.g. what if AI increases productivity growth by 0.4%? Do data centers use too much water? Are AIs racist when reviewing resumes?

The frontier labs, on the other hand, are thinking about replacing all human labor, ending death, and the risk of it causing human extinction. Most of the apparatus we're talking about approach it very parochially; it's almost like they're embarrassed to take the grander ideas even a little seriously, for being too nerdy/sci-fi.

scarmig

há 12 diasdiscuss

Google has a compelling story for many AI scenarios: it has lots of outs. It's the only frontier lab for which that's true. A massive bubble bursting wouldn't be existential for Google; it would be quite painful, but survivable, and even offers some potential upside (picking up assets and researchers from the wrecked, mangled corpses of other frontier labs on the cheap).

scarmig

há 13 diasdiscuss

Just own the wrinkles. Linen isn't meant to look perfectly structured.

scarmig

há 14 diasdiscuss

The latter question is 100% reasonable, and something I also fall on.

But there's definitely a large contingent who denies that they think there's any risk at all, instead of them engaging in motivated reasoning to think their self interest just so happens to coincide with what is best for safety.

scarmig

há 17 diasdiscuss

Useful info, explains a lot. Thanks.

scarmig

há 17 diasdiscuss

Do people in Germany see limited liability as some kind of crazy thing only a scammer would want?

scarmig

há 19 diasdiscuss

Anti-AI psychosis. Anyone treating AI as a serious topic beyond "it's spicy auto complete/obvious scam/stochastic parrots" is some mixture of evil and stupid and can be dismissed out of hand, regardless of their argument.

scarmig

há 19 diasdiscuss

The point here is not the correctness of their beliefs. It's about what the actual content of their beliefs is. If someone says that Donald Trump is a secret Shia supremacist and using that to explain his actions, pointing out that that belief about Trump's belief is wrong is not a statement that Trump's beliefs are correct, but that your model of Trump's belief is incorrect.

scarmig

há 19 diasdiscuss

> This would be a great point if I had introduced either the NFT/crypto comparison OR the "existential threat" parameters, but if you read through the thread, you introduced both.

To quote you, ten minutes before:

> that the apocalypse is coming and your jobs are gone and everything is going to be shitty because a bunch of ultra rich midwits want even more money and there's nothing you can do to stop it.

It's a fair description of your stance. And, full of sound and fury, signifying nothing.

scarmig

há 19 diasdiscuss

Sure; they might be caught in their own hype cycle. I am pointing out that the repeated claims in this thread (and by Geohot, ironically enough) that they are NFT crypto bros running a scam is incorrect.

scarmig

há 19 diasdiscuss

Their usual stance is something along the lines of their company is creating AI correctly--not that AI will inherently destroy humanity--and that them working there helps that end. It's extremely reasonable to question their approach, but here you've jumped from "NFT crypto bros trying to run a scam" to "monsters excited to annihilate the human race," which is a wild leap, and betrays a point of view that seems more driven by anti-AI psychosis than consideration of any evidence.

comments