I've been thinking about the large company money exchanges going on between OpenAI, Microsoft, Nvidia, AMD, and others.. and I think there's a case that the bubble has already bust, we just don't see it reflected in stock pricing or corporate profits just yet. Here's my thinking:
- Nvidia is effectively giving OpenAI (and others) a "rebate" on their GPUs by seeding them money. One source I heard claimed their outward investments equate to about a 25% of the cost of sold high end AI GPU hardware.
- OpenAI is expected to have ~$12B in revenue for 2025, but lost $13B just in the first half of 2025; with spending projected upward significantly for the next 5 years.
- OpenAI's cash position is constantly dwindling; they had "$10B to $20B in cash at end of 2024", external investments are required to keep them afloat.
- Global politics and the trade wars have meant the total addressable market (TAM) is shrinking for companies like OpenAI. At the same time, this is forcing major markets like China and other countries to invest in their own AI.. further reducing the AI TAM potential for OpenAI, Google, Meta, etc. This also means companies like OpenAI are competing against foreign governments..
- Some companies are already starting to give up (smaller AI efforts) and/or slow down AI efforts (Apple).
- OpenAI is starting to exit the "build it and they will come" phase, with features like news being shared with users who log into ChatGPT, based on their past chat history. I interpret this as the growth isn't coming fast enough organically and they're trying to create demand.
- There are physical hard caps on how fast AI can grow regardless of demand: Energy, capacity at TSMC, datacenter capacity (even Oracle's DCs won't come cheap), etc.
..
In summary, I think Nvidia's (and others) investment cycles are merely buying time hoping that a small economic miracle (10X growth in 3 years) will come to justify those investments. The fundamentals of AI revenue aren't looking solid now or near term - as AI is being scaled with the intent to be used "by billions of people" when only millions are finding economic value in it currently.
I recognize this is a bit pessimistic, but I'd appreciate thoughts on this topic.
P.S. I don't own any of these AI companies directly, though I do own broad mutual funds that would be affected by pricing here. I also do not short stocks - I just wanted a disucssion here as I like to learn about fundmentals and imagine what's next. (AI or anything else driving funding for new nodes is exciting to me).
- Nvidia is effectively giving OpenAI (and others) a "rebate" on their GPUs by seeding them money. One source I heard claimed their outward investments equate to about a 25% of the cost of sold high end AI GPU hardware.
- OpenAI is expected to have ~$12B in revenue for 2025, but lost $13B just in the first half of 2025; with spending projected upward significantly for the next 5 years.
- OpenAI's cash position is constantly dwindling; they had "$10B to $20B in cash at end of 2024", external investments are required to keep them afloat.
- Global politics and the trade wars have meant the total addressable market (TAM) is shrinking for companies like OpenAI. At the same time, this is forcing major markets like China and other countries to invest in their own AI.. further reducing the AI TAM potential for OpenAI, Google, Meta, etc. This also means companies like OpenAI are competing against foreign governments..
- Some companies are already starting to give up (smaller AI efforts) and/or slow down AI efforts (Apple).
- OpenAI is starting to exit the "build it and they will come" phase, with features like news being shared with users who log into ChatGPT, based on their past chat history. I interpret this as the growth isn't coming fast enough organically and they're trying to create demand.
- There are physical hard caps on how fast AI can grow regardless of demand: Energy, capacity at TSMC, datacenter capacity (even Oracle's DCs won't come cheap), etc.
..
In summary, I think Nvidia's (and others) investment cycles are merely buying time hoping that a small economic miracle (10X growth in 3 years) will come to justify those investments. The fundamentals of AI revenue aren't looking solid now or near term - as AI is being scaled with the intent to be used "by billions of people" when only millions are finding economic value in it currently.
I recognize this is a bit pessimistic, but I'd appreciate thoughts on this topic.
P.S. I don't own any of these AI companies directly, though I do own broad mutual funds that would be affected by pricing here. I also do not short stocks - I just wanted a disucssion here as I like to learn about fundmentals and imagine what's next. (AI or anything else driving funding for new nodes is exciting to me).