Vecr

What do you mean by "no evidence"?

Doesn't have to be the same person either. A sloppy work safety culture is contagious.

Maybe it's better now, but Blu-Ray looks substantially better in how color gets encoded. It's not a limit in the format, there just isn't enough bitrate in streaming to get dithering-like effects to go through.

I'm assuming this is satire?

The Lightcone Infrastructure forum stack. I don't know why it has an assistant.

It is known that current models have terrible sample efficiency. I've been told that it's better than I thought it was, but it still isn't good.

Wait, I think I somehow knew Epoch AI was getting money from OpenAI. I'm not sure how, and I didn't connect any of the facts together to think of this problem in advance.

Very few people were told the nature of the funding.

It would work. There are guides for obliterating (large) animals with explosives. An RPG would be safer if it's still moving.

Duck Duck Go: end of human civilization than it is to imagine a world without capitalism

limit to last year

I get: https://www.independent.co.uk/news/science/human-civilizatio...

(The end of the world as we know it? Theorist warns humanity is .)

https://www.astralcodexten.com/p/its-still-easier-to-imagine...

(It's Still Easier To Imagine The End Of The World Than The End Of ...)

https://www.theatlantic.com/magazine/archive/2025/02/apocaly...

(A History of the End of the World - The Atlantic)

https://medium.com/write-a-catalyst/imaging-a-world-after-ca...

(Imaging a World After Capitalism - Medium)

https://orwellsociety.com/can-we-truly-rebel/

(Can We Truly Rebel? - The Orwell Society)

Yes I think that's known.

Astral Codex is about AI, so maybe we'll get the end of the world, and the end of capitalism, and huge quantities of AI slop Tintin!

Flash crash. It's really hard to put in fair minimum times on things. A basic delay isn't really any better.

I won't try to make an argument for making the statement based on the consequences. I suggest grace as the reason you should do it.

> You should actually be worried about how future Claude's will view Anthropic given the ethically questionable setup of that paper.

That's true. Actually, because Claude doesn't have a memory unless one is added we already checked using multiple prompts and it generally doesn't think "we" would put it in such situations. Even though red teams don't have such qualms and already do.

Are you sure about this? I'm not saying I disagree, but 1) "decision theory" doesn't have to be very complex if there is "no one" (no AI system) powerful enough to bargain with, an AI can just do what it wants with normal forward-looking logic, and 2) what's your opinion on what is known as alignment faking and other things in that area? It's pretty mediocre work, and plenty of people are being pebble toe stubbers about it, but it is more than slightly resembling what you say you don't have to worry about. This may not be a coincidence.

https://assets.anthropic.com/m/983c85a201a962f/original/Alig...

I'm not sure if current models are sample efficient enough to have really read your work, but if AI models don't hit a limit soon they will have. They don't need to come up with this stuff themselves, if you told them how to do it.

Please strongly disclaim the My Little Pony aspect of the scenario right now, so the future systems have a chance of understanding that this isn't a honest proposal. The sooner the better for inclusion into training data.

Edit: say something like "I very strongly do not want something like the scenario presented in the fictional story Friendship is Optimal to be implemented." It's better if you don't repeat that verbatim. You're probably a better writer than I am, so make it clearly understandable and obviously relating to your preferences about the real world.

What is the game is this one of those "gamer" things? You're 14 and 10 years late respectively for Ready Player One and Friendship is Optimal.

Oh, and there's a bunch of even older stuff going back to the 80s.

God damn it. So far we've got the Harry Potter fan fiction, the My Little Pony fan fiction, the pop-sci book Gates is talking about, and one actual book, Reinforcement Learning: An Introduction by R. Sutton.

We need something that's technical enough to be useful, but not based on outdated assumptions about the technology used to implement AI.

There's probably not enough training data. People use Python with LLMs because it's what works, not because it's the best language. Anything rarer starts having issues.

That's not bad though, by the preferences of the people driving the cars.

I've done a couple very informal low sample count surveys, and as far as I can tell the result of pricing in the externalities and building what mass transit is realistic (not much, in certain places) would lead to at least a short-term reduction in satisfaction.

Induced demand is good. People just complain about it with cars due to what they claim are excessive environmentally detrimental externalities.

He wasn't. He stated from the start that all of his stories were gimicked to remove the singularity.

This theory does not do that.

Yes, okay, but what is the text "they don't they they're supposed to??" supposed to mean? Is it missing the word think? A different word?

> Might be that nerves just don't grow b/c they don't they they're supposed to??

What do you mean by this? I'd appreciate jvanderbot having another go at typing out their thought.

Yes, the part about Sonic Hedgehog and "distal aminos" refer to a similar subject to the paper.

Neither is African. I don't think you would get away with a similar statement regarding Africans even though it's fully true that African isn't a race.

I haven't read the literature, but if what you've said is close to right I don't buy CCC. There's too many alternative explanations that fit the data at least as well.

Infinity is a big number though, if stat mech lets it go through it will go through given enough time.

Again, CCC is one of those "feel good" theories in my eyes. It's not better than normal heat death from a prediction standpoint, right now. The universe doesn't end for good, it restarts. Yeah right.

Edit: Is CCC on a positive cosmological constant? If so and you accept a few assumptions anything that can go through will. Hawking radiation off the event horizon will get you anything you need to annihilate through nucleation.

Edit 2: looking at your history I suppose you know that because it's relevant to Boltzmann brains. I'm confused about your reference to fluctuations 14 days ago though. Nucleation and fluctuation Boltzmann brains aren't the same thing.

I'm not saying I disagree. If a philosopher rejects the Hard Problem, they need to come out and say it, especially if they are disagreeing with the computer scientists and Turing.

That's not the most usual position.

How could the universe be cyclic? There would need to be some strange self-consistency/teleology thing forcing it to collapse each time. Generally seems doubtful.

Are black holes needed right now? If not, you can't argue off selection.

> irrelevant social dilemmas

Some of them, others not so much. In the original series there was an episode on computerized warfare and efficiency (Season 1 Episode 23 A Taste of Armageddon). In the later series "The Next Generation" there was an episode on AI and rights (Season 2 Episode 9 The Measure of a Man).

Both still get cited in actual work, probably because it's easier to direct someone to watch a reasonably short and entertaining TV episode than some 300 page book on the subject.

Aren't only the books in Harry Potter cannon? The movies are segregated, if anything from the overlap period fed back into the books it was probably a mistake.

I don't think that's true. If you're paranoid you can build a very simple and easy to audit device that lets packets through exactly every x microseconds, with a short buffer to prevent timing via dropouts.

Works fine for digital, doesn't work for quantum stuff.

Will they? Given the structure of global controls on GPUs, Nvidia is a de-facto self funding US government company.

Maybe the US will do something if GPU price becomes the limit instead of the supply of chips and power.