Anthropic's S-1 Shows AI Is Already Profitable, So Where's the Bubble?

No one seems to worry about the demand of ai. That makes me wonder. My neighbors for example are normal working ppl they don’t know how to open a zip file (they call me to help them with something like this) and they are in their early 40s so not that old. Sure companies and workflows are going to become agentic but not everyone needs it or even wants it. Then there are smarter small models like Qwen 27b running on a Mac mini and this will also continue to get better and develop further. If the usage is easy I see ppl downloading a desktop app that’s dead simple instead of paying for ChatGPT. I just don’t see this growth curve going for 20 years for these companies to pay off their dept

Railroads, semiconductors, plastics, networking... the utility was there for all of them, and so was the potential profit.

Everybody knew it.

But some investors don't have the patience for these "sure things" to mature into their commercial destiny and others know they can siphon off a killing by taking advantage of them as they speculate.

These bubbles aren't about potential, they're about timing.

And today's AI, for as revolutionary as it may be, is taking just as long to mature as anything in the past.

Most of the top players involved in AI right now are going to be rich. And in 20 years, the world is going to involve a lot of AI in soooo many things. But a lot of shirts are going to be lost below those top players and a lot of companies are going to fold because infrastructure, refinement, and integration take time while the FOMO speculation starts as soon as the big discovery is recognized.

> Profits Can Grow, By A Lot

Profits are mentioned but then only revenue is discussed, Revenue != Profit. Revenue can grow yeah but experimentation/token-maxing stage is ending and price wars have started, so I'm not sure about rising profits, AI Labs won't necessary be able to capture the margin(>>China<<), it might be companies/consumers who will capture most of it, if so there is a bubble - it is in AI labs with their trillion $ valuations and they might end up in a race to the bottom, BigTech is not in a bubble for the most part, they will endure a correction and fall back on their existing business.

Big red flags that I think point the other way:

- anthropic IPO repricing

- pareto frontier gap decreasing by the month between open and closed models

- SpaceX disappoinment

- enterprise buyers as main consumers of closed tokens are being displaced by OS competition for pennies

- job loss and entry level hiring collapse

- massive discounts of current data center contract debt

- consumer open weights model efficiency improving constantly

- the Chinese arms race

- a possible rapid downward spiral that may force governments to step in and subsidize

Am I missing anything?

The entire premise on whether it is a bubble depends on if LLMs can be more than a next token prediction machine or not. Adding reasoning agents, agent swarms, and so on to enhance the results from what is essentially just predicting the next token still means this technology at its core only predicts the next token. It is not AGI and unless something fundamental changes, it never will be. There are certainly valuable applications for this technology as we have seen in coding, math, and other sectors. But it is still not more than taking in a lot data and predicting the next token.

Meanwhile, the major tech companies first financed this build out using the cash flow from their existing revenue. Once this became more difficult to justify, they literally set up companies to keep this off their books directly, and these companies are ultimately designed to take a loss and fail. But before reaching this failure stage and defaulting on debt and obligations, they went to the criminals on Wall Street and designed a financial product structured not all that different from mortgage backed securities. Also, as the "AI" aspects of their existing businesses expand, it eats into the traditional revenue generated by their core business, which if we remember, is the original collateral being used to justify all the debt in the first place. Now these core tech companies have to win "AI", whatever that ultimately means, to not only pay the debt from the build out, but to also replace their ex […]

Im not convinced by the claims of infrastructure only being expensive to build initially. Hardware doesn't last forever especially in a data center, maintenance isn't free, electricity is not going to get significantly cheaper. If models get more efficient we won't need super expensive and expansive hardware build outs, if models keep getting more powerful and needy and are hardware limited, each new hardware generation will stomp over all this expensive and soon to be old hardware in electrical efficiency requiring another massive round of funding that players already holding debt will struggle to afford.

Napkin math: if the 5 trillion investment is paid back by 1 billion customers, every customer pays $5000. If they want to write off the investment in 5 years, that's $1000 per year.

Even with the current AI capabilities, with the right products that seems doable.

I dont know if the valuations make sense, but the usability is all there and so are the potential profits. Over a decade, that's probably all that matters. If the valuations pop or dip (anytime soon), I think that's mostly irrelevant. I get that the press probably doesn't like to hear this boring conclusion.

P.S: A Matlab/Physics simulation guy running for a Liberal party seat in Canada? Say more!

Two things can be true at the same time:

1) There's a huge demand for AI.

2) AI is way too expensive for the companies building it.