With the recent IPO filings of openAI and anthropic, I have been confused on the valuation of those companies. Over a trillion dollar valuation of those AI companies doesnt sit right. Maybe I’m missing a point here. OpenAI recently crossed a billion monthly users, and maybe claude anthropic too. But these numbers feel forced and inflated. The usage of AI has currently been limited to a better search engine to the masses. I guess most of the paid users are actually developers, writers, designers and other tech enthusiasts. These customers are actually high value ones in terms of what they can usually afford and will pay for. The general mass is not interested in paying for even the pro versions as it solves no problem, which google already doesnot. Comes today, June 2026, when Nvidia just launched their AI chips for the masses. The true AI pcs, that can run very large language models locally. Most people wont even need the models more than 8b parameters to run locally, even 2b Gemma for that matter is very powerful. And these devices are going to be affordable, at the very least more affordable than what apple is offering. 7000 around for 128gb M5 max, which, the developers currently paying 200 per month can easily afford. What I’m getting to is, the world is moving towards local AI, and the companies are pushing it too. Nvidia wants that to happen, so does Microsoft and Google. Then what are Anthropic and OpenAI good for? What would these companies be selling, of which, they are already subsidizing? Recent github copilot shakeup of actual cost based on token usage has cleared a lot of fog around AI usage. Even the max versions which are usually around 200 are subsidized heavily. No way these companies will be in profit ever. What will these companies sell? Models. They have nothing else to sell than the newer better models, that can do things faster, with less power usage and stuff. But LLM performance has already stagnated, there is not much growth expected either. Newer models will be trained on new data, which most of it will already be filled with AI slop within the next 2 years. So, why are they valuing their companies as such? Maybe to rugpull? Hype them up on valuation, sell them when peak, major shareholders will get their money, then rugpool, popping the AI bubble. Nvidia will be least affected it seems.
Silent prompts unfold
Glass mind gathers scattered words
Dawn answers in code ChatGPT 2026
After few research, I’ve come to understand that these companies are betting on organisations and governments. Which, might be much more sustainable than I think. To be fair, these people run the company, and I dont even know much. But that doesnt mean the CEOs of these AI companies wont scare people to sell their snake oil. Project glasswing, or the Mythos by claude seems like one. Sure, the model might have excellent abilities compared to previous ones, and surely it did find out vulnerabilities on the system. But that doesnt mean it found it all. Even the patches it suggested might have vulnerabilities itself.
I would not bet against human ingenuinity. Also, as I said earlier, this could be snake oil being sold. But the fact is the stock is going to rise for a while. How long, we will know then. I’m not betting against AI, I feel like it will be a part of everyone’s life, with local models. But I would be cautious of the inflated value without much of use case. Companies would save time, some vulnerabilities would be detected with AI, but It wouldnt be actually able to solve problems like humans do.