www.Tobold.com
Tobold@GMail.com
A blog with my thoughts regarding games I am playing and other stuff in life. Please read my Terms of Service
RSS Feed
Thursday, October 16, 2025
Sounds like an admission of failure
"The internet is for porn" is one of the oldest memes on the web. While many internet businesses managed to attract millions of people, they often had difficulties to then monetize that success to the point of good profitability. The exception has always been porn, where even free porn is surprisingly profitable, and OnlyFans is making billions. So it is probably not surprising that OpenAI wants some of that money.
Of course they aren't the first ones. There are already lots of platforms offering "AI girlfriends", including paid-for NSFW options. But what struck me about the OpenAI announcement wasn't the shock that the internet could be used for porn, but the implied admission of failure: The valuation of all those AI companies is linked to a general idea that AI is going to make hundreds of billions of dollars by replacing a huge number of human jobs cheaper. Even if OpenAI "Erotica" would end up making as much money as OnlyFans, this is hardly in the same league of financial profits. Global AI investment in 2024 was $252 billion, which is about 4 times the size of the global adult entertainment market. You simply can't reach trillions in market capitalization with just porn.
For me this is another sign of the impending bursting of the AI bubble. I am seeing more and more comments in which analysts when asked about the risk of a bubble point out how just before a bubble the stock markets rise the most; sophisticated financial institutions have faster access to the stock market, and can bail out faster than retail customers, who would be left holding the bag when the market crashes.
- posted by Tobold @ 8:08 AM Permanent Link
Comments:
<< Home
Newer› ‹Older
Half of USA GDP growth is from AI. As a lot of experts describes AI as a bubble, if they are correct, it will be a massive crash.
The sign are very bad : each generation of AI is nearly 10x more expensive, with less than linear increase in performance. 95% of employee with access to AI do not find it beneficial. Last year, OpenAI was losing money for each paying 200$/months customer - and obviously from free one. And the huge investment are in computing power, that lose values very fast ( Moore's law + quick deterioration of 'burnt' GPU).
In the internet economy, the wisdom is to lose money at the begining to build a huge customer base, that you can later milk to cover your initial investment. But here, each additionnal customer is decreasing your benefit, and your HW investment will be worthless in 5 years - while competitors can use your AI to cheaply train their own.
So what is the rational for those trillions of investments ? It is a bet that Global intelligence will be so powerful that it will instantly break economy, before anybody can catch you. Half of US GDP growth is a bet on this possibility.
The sign are very bad : each generation of AI is nearly 10x more expensive, with less than linear increase in performance. 95% of employee with access to AI do not find it beneficial. Last year, OpenAI was losing money for each paying 200$/months customer - and obviously from free one. And the huge investment are in computing power, that lose values very fast ( Moore's law + quick deterioration of 'burnt' GPU).
In the internet economy, the wisdom is to lose money at the begining to build a huge customer base, that you can later milk to cover your initial investment. But here, each additionnal customer is decreasing your benefit, and your HW investment will be worthless in 5 years - while competitors can use your AI to cheaply train their own.
So what is the rational for those trillions of investments ? It is a bet that Global intelligence will be so powerful that it will instantly break economy, before anybody can catch you. Half of US GDP growth is a bet on this possibility.
The concerning part to me about all this is not if the AI bubble will pop, because it will, but what happens to all the other systems, like our energy grid, that have rapidly expanded to keep pace with AI investments that will never be profitable.
When AI companies go bust and investment dries up its going to be regular people footing the bill for all the new infrastructure that wont just disappear if OpenAi closes up shop tomorrow.
When AI companies go bust and investment dries up its going to be regular people footing the bill for all the new infrastructure that wont just disappear if OpenAi closes up shop tomorrow.
Youtube sends me an equal number of videos on AI, in which either the bubble is about to burst and we will face an ominous crash, or AI is about to gain general sentience and kill us all by 2027. If one thing has always been true, it is that real life always favors the less exciting pathway, at least when it comes to tech, so I'll just keep monitoring my investments and brace for the inevitable.
I think OpenAI is already all in, and gambling everything that if & when the bubble bursts, they'll have such a lead and be so integrated into everything that they're too big to fail. See 2008. They're approaching a billion *weekly* users already, allegedly. I'd be a lot more nervous about also-rans without a core side business, like, say Anthropic & Perplexity. Google, Meta, and X can probably weather the (considerable) storm.
Recently saw this video that explains fairly well why the AI industry is currently a bubble that will eventually pop.
https://youtu.be/Q0TpWitfxPk
Post a Comment
https://youtu.be/Q0TpWitfxPk
<< Home
