Forgot your password?
typodupeerror

Comment Re:They distilled human knowledge (Score 1) 92

Yeah, I used to do that too. Decided to stop bothering with the quotation marks a couple months ago.

We're not going to spend the rest of our lives putting quotations around words when talking about models. "Think" and "reason" the words we have in English for what is going on. No need to tiptoe around it. Again: models are not humans. They are not the same as us. But those are the words we have in English for what they're doing.

Comment Re:They distilled human knowledge (Score 1) 92

They are, by definition, Markov chains.

Even in your attempt to be pedantic here (in which the universe and everything within it is a Markov chain), no, it's not. The hardware state is Markovian but the linguistic processing is a Nth order autoregressive process; it depends on the N previous states. Also, your argument is akin to saying "a Boeing 747 is just an arrangement of quarks, so don't get hung up on aerodynamics." it entirely ignores the relevant architectural details, and instead substitutes a model that blows up exponentially explodes in size within a small number of states.

If you tried to build a Markov model to do what LLMs do, and could store one probability in every unit of Planck space across every unit of Planck time, it couldn't handle a prompt longer than about 2/3rds of the first sentence to A Tale of Two Cities.

Comment At that price, yes, an experiment (Score 1) 15

Some people will certainly buy it at that price. I'm not even saying it's overpriced for what it is today. I'm saying you can't make it highly successful at that price.

If they can cover their costs they get into a product and get a bunch of experience and IP for the future in which it actually is viable to make it cheap enough to be successful, so it might be worth doing. But it's hard enough to sell the kind of game consoles everyone is comfortable with at that price "in this economy" etc etc.

Comment Re: Cost (Score 1) 78

Here in the USA I have seen both types of behavior from unions. That is, I have seen people legitimately protected, and I have seen fuck-offs who literally could not do their job protected from being treated appropriately. I was literally contracted to do parts of his job, doing things I had never done before, read the docs, got them done. He was supposed to retire and I probably would have got his job, but he bought another motorcycle (had two already) and had to keep pretending to work in order to pay for it.

Unions can be good, but unions can also be shit

Comment Re: Renewable+storage = reliable (Score 1) 115

PGE didn't have any clue how much most of their lines could carry until recently when they hung temperature sensors on them. You can bet there are still a lot of parts of the line which are unmonitored because PGE is a gigantic shitfest. I thought you knew stuff about power distribution, or you thought so anyway, why didn't you know this?

Comment Re: Bullshit. (Score 1) 114

Can't do it. I'm refactoring on a large code base. I eventually had to tell it to work on just one task and ignore other concerns it had to prevent it from investigating until it got stuck in a cycle of investigation.

Understood, and yeah single tasking seems like a good workaround given the situation. Or switching to an LLM with a larger context window, if possible/practical. I was able to get about 7X more context window "runway" using the AST approach, but it's a relatively small code base to begin with.

Slashdot Top Deals

Don't steal; thou'lt never thus compete successfully in business. Cheat. -- Ambrose Bierce

Working...