Hacker News Reader

Comment by rich_sasha | Hacker News Reader

I largely agree with this take. What gives me disquiet is that I thought more or less exactly this about Uber, and was proven comprehensively wrong. Despite burning money like a furnace on an app for taxis, it seems that anyone who invested privately made a handsome return, and they are at a profitable steady state. Is AI the same?

View on HN

The cash burn is not the problem. You can’t build a big business without it. The difference is network effects. With ride sharing apps, you need a lot of drivers and riders collected on one platform. It is nearly always better, for both drivers and riders, to switch to a larger platform. Thus, it was worthwhile to spend the money to become the biggest fish. Once this was accomplished, Uber could raise prices because switching to a smaller rival would mean less availability, and thus less utility (for riders) or less earnings (for drivers).

With AI, by contrast - at least in its current state - there is no benefit to be gained from using the same model provider as somebody else. Switching is trivial for most use cases. Since they can’t capture consumers using network effects, the labs only have the levers of price and quality to pull to acquire and retain customers. To pull the price lever, they have to reduce their revenues; to pull the quality lever, they have to increase their expenditures. Indeed, they are sowing the seeds of their own demise by making inference cheaper and more efficient: since they can’t exercise pricing pressure, efficiency gains will be passed on to the consumer, which is unsustainable if your GPU debt is priced based on yesterday’s efficiency expectations.

This only works up to a point. Plenty of different makers of phones for example all making fungible phones at ever lower prices, yet managing to eke out a profit. I can see a world where OpenAI, Anthropic, Chinese companies corner the market, make themselves indispensable and start charging market rates, while also getting better and more cost efficient.

Here again you have network effects, though. I personally have an iPhone because it’s what all my friends and family use and there are communication functions that are much easier to engage with if you all have the same sort of device. It’s also not trivial to switch, you need to, at minimum, go to the store or wait for something to get delivered - to say nothing of the wasted money from buying multiple phones. If I have unused OpenAI tokens I can burn them on side quests or something. If you have an unused iPhone you need to find a way to sell/return it or eat the cost. There’s also potential violations of your contract with your mobile carrier, etc.

If I’m using ChatGPT and I decide I want to use DeepSeek instead, I am only a couple of keystrokes away from doing it, and that’s if I have never used DeepSeek before.

As for your “corner the market” scenario, it’s possible, but unlikely. It is too easy to enter; even if you somehow got all of the major players to commit to growing their margins - and somehow manage not to violate the antitrust laws in the process - a newcomer could spoil the party far easier than it could in an industry like mobile phones (where you need tons of components, manufacturing capacity, network relationships, etc.) or ride sharing (where you need a large user base to justify your existence).

Mmaybe- not sure. For technical applications like coding, perhaps. Though even there you have plenty of opinions about different models’ differences on subtle features, this model writes bad tests, that one is bad at JavaScript etc.

But the real sales pitch is that AI overtakes everything. That the YC cohort of 2032 will be just CEO, sales guy and a massive AI bill. It’s not entirely impossible IMO, either.

Then you totally get network effects. All your company documents, discussions, context etc are in there, your agents/employees whom you finally taught to do the job right. And I’m guessing the REST API for extracting your data is absent.

I don't know.

But the economics of it, at least from the outside, smell funny.

Some people are anxious that AI will take people's jobs. I'm not. I say this as a daily user. LLMs can be very useful, but they need to be carefully steered. It feels like a superpower the more I am an expert on the subject matter. What I am afraid of is that I suspect that once the dominoes start to fall, the economic downturn that it will spawn will be very, very painful.

Zitron is a bit histrionic, and this may put off people that don't like his style.

If you want a different, more balanced analysis, let me recommend you this: https://youtu.be/NufJ7g63KSY?is=Ojgb5pzrI-wg9wbo

Patrick Boyle's more recent video goes from a different angle and was very interesting for me, that have only a passing, layman's understanding of corporate accounting and investments.

Unfortunately I agree and I think we are going to be trapped in a double whammy scenario: genAI is good enough to justify significant layoffs without being profitable enough to justify the investment. The pain on the stock market will cause employers to adopt AI to shrink their workforces even more than they might have otherwise.

Taxis in general are a working and proven business model. You could have been skeptical about the way they approached it but surprise surprise, you can make money providing taxi services. Less so with "AI", it's not proven at all that this can be made profitable in the near future.

A lot of smart people said that Uber could never become profitable because the core economics couldn't work out:

> https://www.bbc.com/news/technology-48227381

> https://www.forbes.com/sites/lensherman/2019/08/22/ubers-dub...

> https://americanaffairsjournal.org/2019/05/ubers-path-of-des...

Cory Doctorow, who is close to Ed Zitron and writes a lot in the same way about AI's economics, was of the same opinion: https://doctorow.medium.com/no-ubers-still-not-profitable-2b...

I used to believe this, so I'm not sure what to think of the AI market.

The failures of past doomsayers are important to keep in mind. It’s hard to say “well, but we know things they didn’t know at the time, but this time that won’t happen.” I will observe, however, that Uber’s revenues come from consumers. Consumers are never going to be willing to spend enough on AI to justify OAI and Anthropic’s spending, even if a single company captured all of their demand. They recognized this and pivoted to software engineering and enterprise customers. These are much more sophisticated buyers than someone trying to catch a cab or get a burrito delivered, and more able to be patient or expend resources to find a better, cheaper provider.

It also bears mentioning that if any particular AI lab manages to survive and succeed in the way Uber has, there will be several multibillion dollar corporate gravestones behind it. In fact, I don’t even think that nobody can be the Uber of AI. I just think it can’t be OAI or Anthropic. The debt is too great and it’s priced under old assumptions.

Reply on news.ycombinator.com