Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

I love LLMs, I hate hype

geohot.github.io

Post reply on HN
Print view View original

141–150 of 340 posts

Re: I love LLMs, I hate hype

#141

Post by rglover »

deep_concern wrote: post #69

>One, this constant bullshit about some window closing, or the perpetual underclass, or falling hopelessly behind. This is negative valence hype, not only is it not true, it’s mostly designed to make you feel bad about yourself and move to shitty San Francisco where everything really does suck like how these people claim. It's bullshit in the sense that they don't know for sure, but the author doesn't either. Why mig…

May be true because humans, especially in the West, are big on performative humanitarianism but not actually considering the well-being of others (or changing their behavior solely for the benefit of others).

May not be true because it's a blind spot to assume that purely by being a player in the AI game (with no real attention paid to quality of result), you have increased odds of winning the game. That's true in the abstract, but practically, it requires a competent player to become true in reality.

Re: I love LLMs, I hate hype

#142

Post by sixtyj »

arjie wrote: post #113
sixtyj wrote: post #86

Earlier quoted context omitted.

Creating a fork of an active project only makes sense if you are its sole user (of the fork) and you really need exactly the modification you've been dreaming of. I have seen so many unnecessary forks of popular projects that I think it's better to stick with the original, even if that means it won't be perfect.

In the old world, this was because keeping in sync with upstream was hard. In the new world, it takes an hour. And because you're the only user, you can test in prod. Makes the whole thing faster. I have lots of forked and family-only software. Some are abandoned upstream etc. As cost to software goes to zero, these things become easily possible. In the past, I'd only fork top-quality software (things like `xsv` etc.…

The law of conservation of energy also applies to software. If the price of software approaches zero, it is offset by the time and tokens required to modify and maintain it. Price, time, tokens are simply different expressions of energy.

Re: I love LLMs, I hate hype

#143

Post by SwellJoe »

bluegatty wrote: post #114
SwellJoe wrote: post #105

Earlier quoted context omitted.

When do you reckon they'll start doing that?

They are raking in 10's of billions, growing revenues astronomically as we speak. That's not sign of commodity actor, just the opposite.

But, that's not what you said they'd do. I can switch to a different model with almost zero cost. That's the definition of a commodity.

Re: I love LLMs, I hate hype

#144

Post by maxgashkov »

QuantumGood wrote: post #95
SwellJoe wrote: post #81

This line: "this is my main argument against the valuation of frontier labs. It’s not that AI won’t create that much value, it’s that they won’t capture it." That is a very astute and concise way to explain everything about how the frontier labs are behaving and how they're trying to push more people to pay token rates for the best models. At the current subscription prices ($100 or $200 a month for a generous, thoug…

Just to clarify your implication: Fable subscription usage was just (re)extended to July 19

If they haven't relaxed the classifier this makes little to no difference as the model is essentially unusable no matter the token quotas.

Re: I love LLMs, I hate hype

#145

Post by Giefo6ah »

andy99 wrote: post #121
SwellJoe wrote: post #81

This line: "this is my main argument against the valuation of frontier labs. It’s not that AI won’t create that much value, it’s that they won’t capture it." That is a very astute and concise way to explain everything about how the frontier labs are behaving and how they're trying to push more people to pay token rates for the best models. At the current subscription prices ($100 or $200 a month for a generous, thoug…

The funny thing about Fable, is we all but know it will be obsolete in a month or two. Between their embargo shenanigans (which IMO they could have avoided just by not pretending it was dangerous) and continuing to give access, whatever marginal advantage it had was essentially wasted. It would have been an interesting experiment to charge more for it right away and see what the market would bear, rather than tease i…

They clearly didn't have the compute capacity ready to release Fable at any price that would have seemed "worth it".

Re: I love LLMs, I hate hype

#146

Post by fc417fc802 »

andy99 wrote: post #121
SwellJoe wrote: post #81

This line: "this is my main argument against the valuation of frontier labs. It’s not that AI won’t create that much value, it’s that they won’t capture it." That is a very astute and concise way to explain everything about how the frontier labs are behaving and how they're trying to push more people to pay token rates for the best models. At the current subscription prices ($100 or $200 a month for a generous, thoug…

The funny thing about Fable, is we all but know it will be obsolete in a month or two. Between their embargo shenanigans (which IMO they could have avoided just by not pretending it was dangerous) and continuing to give access, whatever marginal advantage it had was essentially wasted. It would have been an interesting experiment to charge more for it right away and see what the market would bear, rather than tease i…

I don't see how that matters? It's a treadmill. Sure Fable-v1 will be obsolete but the shiny new Fable-v2 won't. Don't view their shenanigans as regarding Fable but rather their current cutting edge model.

Re: I love LLMs, I hate hype

#147

Post by fc417fc802 »

SwellJoe wrote: post #143
bluegatty wrote: post #114

Earlier quoted context omitted.

They are raking in 10's of billions, growing revenues astronomically as we speak. That's not sign of commodity actor, just the opposite.

But, that's not what you said they'd do. I can switch to a different model with almost zero cost. That's the definition of a commodity.

They can do both things at once (and appear to be doing so).

Re: I love LLMs, I hate hype

#148

Post by fc417fc802 »

InvertedRhodium wrote: post #126
dakolli wrote: post #104

Earlier quoted context omitted.

The fact that you're even saying this it is probably an admission that you do think it's making you dumber. Most people I know, who are honest with themselves, have admitted to me that they feel like it's making them dumber or "zombifying" them. This is also well studied already, https://arxiv.org/abs/2506.08872 LLMs are poison for the brain, I'm almost certain of it, at least when used in the way most people are usi…

Socrates thought the same of reading and writing, that it would weaken the memory and isolate people from one another. 1966 saw the peak of calculator protests, where math teachers claimed similar things of calculators.

TBF I suspect both of those cases are true. Just that the benefits ultimately far outweighed the costs.

Re: I love LLMs, I hate hype

#149

Post by roncesvalles »

SwellJoe wrote: post #81

This line: "this is my main argument against the valuation of frontier labs. It’s not that AI won’t create that much value, it’s that they won’t capture it." That is a very astute and concise way to explain everything about how the frontier labs are behaving and how they're trying to push more people to pay token rates for the best models. At the current subscription prices ($100 or $200 a month for a generous, thoug…

>5.6 Sol, competitive with Fable (or close enough), and it's available via subscription (even the $20 subscription!)

It's not comparable because OpenAI caps thinking to High in the ChatGPT "Chat" interface (and the "Work" thing where it actually does let us use Extra or Max is fucking shit).

GPT 5.6 Sol (High) is almost certainly worse than Opus 4.8 (Extra), and nowhere close to Fable (Extra).

I literally got a refund for my $20 OpenAI subscription after playing around with 5.6 Sol for a couple of hours (yes even on Codex) because it's so unusable and I'd rather just use Fable today and 4.8 Extra starting tomorrow, still within my $20 Anthropic plan. And I'm not even poor.

Re: I love LLMs, I hate hype

#150

Post by stavros »

edot wrote: post #109
Lt_Riza_Hawkeye wrote: post #96

Earlier quoted context omitted.

While I do agree there will be disruption we haven't seen yet, my company is already spending >$40k/day for a "frontier model", so who knows. Then again, they're not using that for coding

What are they / you using it for in such quantities?

I know that some companies are spending around that much per day per engineer. They just want to go as fast as possible.
Post reply on HN