phailhaus

I don't know what you think I said, you should read the context again. This has nothing to do with movies being made with AI.

Altman thinks that movies "will become videogames". As in, he sees media as some sort of ranked ladder where everyone wants to climb as high as possible. He does not understand what movies are or why people like them.

No it doesn't, not even close. You are looking at a video represented as pixels changing on a flat screen. None of it actually exists. You exist, the computer or phone in front of you is composed of mass and energy, and absolutely none of it is algorithmic. World models, as complex as they are, are still just pushing numbers around on paper.

It boggles the mind how these people can be so dumb. He literally does not understand why people enjoy movies.

> as an extreme measure the wealthy are able to take, or desperate indentured servants are shipped off to work.

This is the part I don't think people appreciate: space sucks. It's horrible. There is not a chance that the rich decide that they'd do that rather than lounge on a yacht. The working class goes first into the mines.

It's okay as long as they call it out, but the author doesn't and instead states it like a fact. They could have easily caveated it with "but a magical camera that can see magnetic and gravitational fields" or something. The analogy is fine, but it's not presented as one, so it's not clear how far the analogy actually goes. So the takeaway is "wow technically it's possible!"

Yes, the rest of the article is pretty good.

> Holography implies that you can do that — reconstruct every fold, vessel, and neuron in three dimensions — without actually looking inside. Simply photographing the surface of the brain somehow suffices.

Gah I am not a fan of pop-sci simplifications like this for the sake of a soundbite. The author goes onto explain how no, a picture is not enough, you need to actually measure the distortion of spacetime at the surface. Somebody is going to walk away thinking this stuff is possible, that maybe psychics really can detect something wrong with them.

Yeah and LLMs can't do anything, they can only produce text. These "frontier labs" are looping that with a harness that performs actions requested by the LLM. They are literally saying "we ran a script that hacked you, oopsie!!"

This is delightful! What a nice and wholesome resource.

The story is that the writer is Orcaset, which is trying to sell you their own analysis tools.

And yet, the React/SSR infra has never gone down.

No they don't, human intelligence has the ability to form an internal model of itself, which allows it to "notice" its own mistakes and change.

We used to think that if a computer could play chess, it would be intelligent. Maybe back then they were also people saying "stop trying to draw lines, just admit it's intelligent!!" Good thing we didn't listen to them.

The act of distinguishing between human intelligence and LLMs is what allows us to figure out how to make it better. There are still some deep limitations, and to ignore them is a mistake. That doesn't take away from how crazy good they are.

Yeah that's the problem. Imagine trying to learn calculus as a top-down isometric 2D animated simulation. It doesn't make sense for the topic.

The fact the author thinks every topic can be fit onto a rollercoaster tycoon-style analogy leads me to think they do not actually understand these topics very well.

That's because you're trying to snarkily jump into a conversation you don't understand.

User `flanked-evergl` is asking why do you need a whole new format for this instead of just using json. And the answer is because everyone needs to agree on how to specify the format in the json, which `crote` suggests is as simple as doing `{"schema": ..., "data": ...}`.

There's your format. If you want everyone to agree on the `{"schema": ..., "data": ...}` pattern, you can no longer accept just `[]` as a return type. Every response has to include the `schema` and `data` keys.

> So your argument is that no solution could satisfy your new requirement

No, read my comment again. They could do it but it would absolutely destroy the responsiveness of the app.

Are you serious? Do you honestly think that I am talking about a schema for passing empty lists around?

If you are passing around lists of structured objects, I need you to send me what those objects are supposed to look like even if the list is empty sometimes. If you say "ok so sometimes I'll give you an object with a schema key and a list of data, other times I will just give you an empty list because I don't feel like it", everyone will hate you.

And there you go, just `[]` is no longer valid, so you need a new format.

My new requirement is NOT physically impossible. I did not mean "immediately" as in "zero milliseconds", it meant strong read-after-write consistency. It makes facebook so slow and unscalable as to be unusable. Facebook deliberately decided that stale reads are worth it for responsiveness and scale.

Completely untrue. Every design can be destroyed by a single poorly-thought out requirement. I can take down Facebook's design just by saying "any edit must immediately be visible to all users".

"Quick scan"? It's a few paragraphs. You can't read a few paragraphs? What's the point of replying when you don't even know what you're replying to?

Your theory is that he found "very dark stuff" that is accessible to any MIT student? Schwartz didn't "hack" anything, he connected to their student network and downloaded articles that they had access to but the public didn't.

Nobody expects the scroll behavior to be different on a webpage, so it's always a surprise when a page does something like this. Then, you have to learn to adapt to how it expects you to scroll so that you don't screw up its animations or overshoot.

It's the kind of UI that works if you scroll a specific way. A lot of people probably do. But if you don't, it's all very tedious. I consider that to be poor design.

Look at the HN comments and you'll see many saying they don't get the point

The broader point is that percentages can be misleading, and are often because of that. It makes things sound better. But usually, the more accurate thing to do is use odds-notation ("1 in 50" instead of 98%). Percentages have a kind of singularity at the edges, where small numerical changes have massive real effects. Going from a success rate of 98% to 99% doesn't sound like much, but that's failing 1 in 50 vs 1 in 100. You've doubled the efficacy.

There is simple correctness but there are also second order effects to consider. How does this particular implementation allow you to grow, and in which directions? What does it prevent? If you don't already have an opinion about this, then the LLM is going to do something and you're going to have to live with it, because it has no idea that it is "making a decision". And now, neither do you!

This is why LLMs do their best work at "leaf nodes", building on existing infrastructure but not designing new patterns on their own.

LLMs can't introspect, reason, or build internal models of the world. You can get very far without that, but there are some subtle ways it will bite you, and it's a fundamental limitation. Hallucinations are one: they are the feature, not a bug.

This is why I get worried about the hatred for income taxes. No, you want the government to want you to be richer. If all they raised was property taxes, then they don't care if you get priced out of your home, as long as someone's buying.

The setup may be simple, but what about the maintenance?

That was the fearmongering, which made no sense because advertisers can't put a dollar value on "the AI will kind of sort of mention you", and because every conversation needs an ad. If ChatGPT always snuck in a brand mention even on the simplest questions, everyone would hate it.

Ad technology is really old. They're just going to use the same proven tech that has a track record of creating billionaires: intersperse content with sponsored blocks.