Hacker News Reader

> that is growing

It seems like user numbers are stagnating, and the ad play isn't working out so far. Where is the revenue growth coming from? I don't think API can be the answer, because API has absolutely no switching costs.

> The obvious way to recoup spend is to grow, rugpull by cranking up costs 6x

Presumably, that was the plan, but I don't see how that can possibly work with how quickly the Chinese models caught up.

revenue is probably the better indicator of whether revenue is growing or not

These companies don't disclose revenue.

> API has absolutely no switching costs.

Were that true, many embedded maps would have switched to OSM instead of Google Maps after Google massively increased costs. Adding your business to OSM is free, and API costs would fall pretty dramatically

I'm talking about inference APIs.

> The actual risk that the author does not even broach upon for investors... the thing that will actually torpedo this massive investment are the open source open weight chinese models that commoditize the entire endeavor.

So much this. I've never touched the Chinese models (no particular reason) but it is having an impact as the frontier companies are pushing the price down as a defensive measure. Is this pulling forward what would have happened eventually? No idea.

It is unclear how effective anyone beyond China and Mistral have been at developing cheaper, capable models. It is an expensive business. I'd be curious if anyone had any thoughts on that

Poolside's stuff (US) is pretty good.

> I've never touched the Chinese models (no particular reason) but it is having an impact as the frontier companies are pushing the price down as a defensive measure

I warn you that it is addictive.

I am using DS and MiMo on Pi.dev, and I don't see myself going back to OpenAI. They are criminally cheap in a way that I don't care to spend tokens. That leaves me space to experiment.

And in terms of capability, well... I use Sonnet and Opus at work (provided by my employer), and I see no difference in terms of what I can achieve. Well, besides Claude costing dozens of times more.

> The actual risk that the author does not even broach upon for investors... the thing that will actually torpedo this massive investment are the open source open weight chinese models.

The monopoly will likely then shift from the model to the compute, i.e. who has the GPUs to serve inference at scale from the open weight models. The cloud compute giants have basically bought everything that Nvidia, Broadcom etc. have to offer. Currently, the inference margins are shared between the cloud giants and OpenAI/Anthropic. But if training great models becomes easier for some reason, the cloud giants benefit. Then they'll have used the OpenAI/Anthropic revenue and spending commitments to grow their cloud business, and then can serve other models and make even more money.

Given that OpenAI and Anthropic are private, I don't think there is any risk to retail investors in this scenario. AI not turning out to be so useful, and OpenAI/Anthropic not being able to pay their bills is the correct failure scenario i think, as identified by the author.

The hyperscalers already had the compute monopoly. They just spent a bunch of money on even more compute. Compute is OK because it's reasonably general purpose to reallocate for what comes after chatbots (e.g. consumer robotics which is reasonably likely to take off in the next 4 years).

What kind of robots are we talking about? What will robots do for me five years from today?

Literally everything chatbots don't do along with everything chatbots currently do. They will gradually go from doing your dishes, to contructing an ADU in your backyard by 2040

System 2 (thoughtful slow planning) has already been cracked with VLA models. The seemingly easier system 1 (fast reactive) will be figured out in months... maybe with JEPA

> The actual risk that the author does not even broach upon for investors... the thing that will actually torpedo this massive investment are the open source open weight chinese models that commoditize the entire endeavor.

I disagree that "the entire endeavor" is inherently doomed for Anthropic and OpenAI. There will always be a very top end of the market running humongous models (like the recently-teased OpenAI Astra and perhaps including future versions of the existing Claude Mythos) that's too large-scale to be successfully commoditized, and that's exactly where the ongoing investments in AI datacenter compute and model training are most likely to pay off at some point. Video generation is another emerging AI area that seems to require large-scale compute, though the value proposition is definitely iffy there.

There may be a market for the super models, but the reality is, the super models and the merely ok models already saturate on lower levels tasks.

You don't need Fields math PhDs to make you a web app or plan your vacation.

The remaining frontiers for high value ROI are ultra long horizon tasks, and lower cost faster inference. The former will rely on better architectures which will likely also get commoditized. The latter will probably be led by Chinese companies which already own most of the electronics manufacturing supply chain.

Video generation was already tried by OpenAI and was enormous bust. It busted so hard the current leading proprietary videogen model is Chinese, and the best open weight videogen model is also Chinese.

> The actual risk that the author does not even broach upon for investors... the thing that will actually torpedo this massive investment are the open source open weight chinese models that commoditize the entire endeavor.

Not in this specific article but Ed Zitron has been talking about open models quite a lot

> The actual risk that the author does not even broach upon for investors... the thing that will actually torpedo this massive investment are the open source open weight chinese models that commoditize the entire endeavor.

OpenAI and Anthropic investors yes, however open weight models are good for cloud providers. They can turn the two large customers into direct ai services that can be spread across many customers and reduce the cloud providers overhead on ai services.

Which makes me cynically believe that all this capital being burned that has made memory, storage and GPUs in some cases 5x is a feature, not a bug. "Can't use open weight models if you can't find or afford the hardware to run them on!"

"And having two whale customers is not really out of the ordinary for any software company... it's just the scale that is staggering."

Does the Apple lawsuit over trade scret theft, including secrets concerning "metal-finishing finishing process", suggest that this "software company" may have plans to sell hardware

*secret

>The actual risk that the author does not even broach upon for investors... the thing that will actually torpedo this massive investment are the open source open weight chinese models that commoditize the entire endeavor. If you don't have a monopoly, you cannot rugpull and 6x the costs on the consumer.

This is the thing: let's just reason it out. What if OpenAI and Anthropic both fail and end up in bankruptcy? What is the impact to the hyperscalers?

...they just start selling open weight model inference to businesses, or whatever other entity picks up the scraps from the collapsed labs.

It's abundantly evident that there's major demand for compute, and that it isn't going anywhere.

I think the nuance that's being missed here is that the profit margin of inference serving when there's perfect competition is close to 0.

Zero profit means no recouping investment costs effectively bursting the bubble.

Technological progress can come at extreme pace but the lack of profits can still torch both the startup frontier labs and the hyperscalers.

We plebians can then pick up H100 gpus off Ebay to help us invent new superconductors and generate waifu videos in pur basement...

Profit margin determines whether OpenAI and Anthropic survive. The hyperscalers would be fine with a model where they make huge investments in infrastructure and you pay to access them per use/per hour. That is already their model for cloud compute and it is highly profitable, even though the vast majority of what people do with it (storage, VMs, etc.) is a commodity.

> The hyperscalers would be fine... their model for cloud compute... is highly profitable.

"fine" and "highly profitable" are doing a whole lot of heavy lifting here, actually masking the quantitative reality that these "high profits" are nowhere near the height required to pay back their debts and remain profitable enough to prevent a sharp drop of their stock prices - that kind of development is the opposite of "fine" as far as investors, ordinary people and the wider economy are conserned.

> Profit margin determines whether OpenAI and Anthropic survive. The hyperscalers would be fine.

That's not how the market works, if OAI or Anthro default, they'll bring down the whole market, it'll be 2008 on steroids. Assuming the fire can be contained to only two trees misses the reality of how dense the forest is and how hot and strong are the winds blowing towards it.

I meant that the current business model of the hyperscalers is highly profitable, and that model is selling compute that you could build yourself but you'd rather not to. If OpenAI or Anthropic default because people prefer open source models, they will still have compute for rent. The only scenario in which they are truly screwed is if demand for AI goes down.

> 200 billion dollars a year on a 30 billion dollar a year business that is growing does not really sound as bad as the author makes it out to be

How do you know it's not "as bad"? Whenever things got that bad a crash followed, a lot of people lost their shirts and savings. For them it's that bad and then some. Why mislead these people?

> The obvious way to recoup spend is to grow, rugpull by cranking up costs 6x and reducing inference cost by half... then simply fire software engineers.

OK, I didn't know that brutal/shmutal was the only way to do business these days. If I could only get with the program, I would understand how rosy-smelly the situation is and not at all "as bad as the author makes it out to be".

> The thing that will actually torpedo this massive investment are the open source open weight chinese models that commoditize the entire endeavor.

Oh, wait, I thought it wasn't that bad? Which way is it? Brutal/shmutal failed to work? But, but but, you said, you promised... "not as bad".

Besides, Chinese competition isn't the only way to commoditize AI, new tech developments are the single most significant risk in tech - a well known fact.

> If you don't have a monopoly, you cannot rugpull and 6x the costs on the consumer.

Anyone who invests while convinced that only a monopoly would save his investments has got his brain screwed on backwards,

>Anyone who invests while convinced that only a monopoly would save his investments has got his brain screwed on backwards

Whats the common trait between Warren Buffet and Peter Thiel? They only invest in companies that are monopolies or on their way to being monopolies. As Thiel (who very much has his head screwed on backwards) says "Competition is for losers"

What does Rockerfeller, Carnegie, Gates, and Musk have in common? They own monopolies.

So yes! The only way these companies are going to make their investments back is by being monopolies.

> Warren Buffet and Peter Thiel...

Past Performance Is No Guarantee of Future Results

> "Competition is for losers"

That's how we got where we are, losing the global competition game, high polarization, inflation, debt, and wars. Communist China managed to surge ahead mainly due to their purposeful development of a highly competitive industry and market. Oh, the irony.

> The only way these companies are going to make their investments back is by being monopolies.

Well, they aren't and they won't be, the game has changed. Seems like when you say "monopoly" you mean "on the US market" but that's not enough to sustain anything resembling a good life here. You may be thinking of an isolated and self-sufficient national economy but that's a mirage. The world is still global and only a major extinction event, like an all out world war, can change that - do you want to go there?

Buffet, Thiel, Gates, Musk, etc have nice bunkers to hide in... you don't.

unironically sama is probably one of the most honest players here after all, he is burning money and if/when they have achieved AGI they will ask it in how to make money. its a different investor incentive story than trying to monetize into profitability right now that is indeed doomed to fail against china. Having no idea but a vision achieved is rrquired to be met to have those returns is actually the honest part here.

If Altman seriously believes his role is to create AGI then ask how to make the company profitable, he should get removed from his position of leadership asap. That’s crazy territory.

It’s a grift, “AGI will save us” is the same as Musk’s “Mars colony”, it’s not supposed to ever happen, it’s supposed to be a goal post they ever move further

He said the following in 2019, at a StrictlyVC event, “The honest answer is we have no idea, we have never made any revenue, we have no current plans to make revenue, we have no idea how we may one day generate revenue. We have made a soft promise to investors that once we've built this sort of generally intelligent system, basically we will ask it to figure out a way to generate an investment return for you.

It sounds like an episode of Silicon Valley it really does I get it, you can laugh, it's all right, but it is what I actually believe is going to happen”

Whether or not that’s his current view, who knows, but everyone who’s invested in OpenAI since should have known that this was at least part of his mental model.

https://youtu.be/gjQUCpeJG1Y?is=--fubVyvQnC8rgse

Yeah, though that was before the shift to a for profit organization

>It’s a grift, “AGI will save us” ... goal post...

It's not really like that, at least the fundamentals, not necessarily what Altman / Musk say.

The significant point is when AI/robots can do what we do without us and improve themselves even if humans disappear. That'll be a new era on Earth.