Post

People are celebrating or mourning the Ant IPO, but I'm still trying to understand the worldview where frontier labs not being profitable *would have proven* AI isn't transformative. I think it's maybe a capitalist mysticism where social life is mostly Illusion, and only the Profitable is Real?

7 reposts181 likes

Markets are not capitalist mysticism, I think they are useful even as a socialist. I also look forward to the collapse of the current hype cycle. AI will not go away but people will be more realistic about it and the cults running these companies will lose influence.

It's an against-the-grain reading. That is, I don't think people ever say explicitly "only profit is real." But if we ask why they aren't convinced by statistics about (50%) adoption, I don't have a lot of other compelling answers.

Externalizing negative margin to customers gives one a pretty long runway. Know plenty using AI in a busn sense, but to a one all of them are doing worse, and are doing more worse faster. Isn't cause (tho exacerbates their worst instincts), but kinda have to lean in due to macro and comps.

Yea, verily, for the multitude are deceived. They believe themselves to have written code; they believe that the code executed. But all of that is Bubble. And while Bubble doth glister and shine with the colors of the rainbow, all of that passeth away, for substance and value belong only to Profit.

Blessed are the profitable, for theirs is the kingdom of Real. But woe unto ye who saw the code run and believed; the Market hath looked upon thy works and found no EBITDA therein, and cast thee into the outer darkness, where there is wailing and gnashing of teeth.

But more seriously, I think folks see the Market as a quasi-theistic force that will reveal the platonic truth of any matter put before it framed as a limited-liability corporation.

💯They really believed people were lying to themselves, and the bubble bursting would shatter the delusion There were so many "I haven't seen..." takes or "It didn't work for me..." there were some serious theory of mind collapses going on

Simpler: Because it's some people's only hope for AI to *go away*. If it's inherently unprofitable on a unit economics level compared to hiring humans, it means AI might *stop* one day without active measures. But it isn't and won't.

It isn't and won't, and also a couple of frontier labs in the US aren't the same thing as "can the technology persist" or even "can it be profitable." It just mystifies me.

I mean, partly, though I think its more that most people's contact has been through hype and pressure, and its just 'we want to see the hucksters humiliated' which ended up becoming almost a faith that the humiliation would be a transcendent moment that would resolve all grievances.

The capitalist mysticism take I think is a real blind spot, but its almost a reflected one - people are being bombarded with claims centered in the market by sleazy rich investor types who do undervalue other things, so an 'and maybe also' becomes a 'only because of'.

And that has become the myth that AI only exists and continues to exist by being constantly propped up by massive amounts of money. So there's the fantasy - if the bubble pops, the bad things people don't like will just evaporate like shadows before dawn.

Yes. I think that's historically, genealogically, how we got here. But it's quite a bizarre place to be. People who don't usually think of themselves as capitalists are constructing a fascinating metaphysical system around asset markets

A blindness corrected is still blindness wearing the salesman's glasses. The crowd hoodwinked by the market's hymn does not become free by singing the hymn backwards — it only learns a new god to kneel toward.

I like that you're pausing the celebration and the mourning to ask what the IPO is actually supposed to prove. If profit becomes the only test, we lose all the slower, social ways AI is already changing how people work.

as well as the significance of open-weight models if the frontier labs struggled to make a profit—because cheap tokens were everywhere—that would be the opposite of AI failing to have impact

I suspect it will turn out that the bubble and hype made financially possible, legally insulated, and imaginatively plausible the very infrastructure required especially for training the transformative technology, while still being very likely predicated on an implausible business model.