As bullish as I am about AI, I am unconvinced that OpenAI will make it. On the investment committees that I sit on or chair, I have pushed back on seeking to enter their funding rounds.
OpenAI is valued at $US500 billion according to reports of an $8 billion sale of shares by employees. $500 bn is 40x its annual sales run rate of $12 billion.
The buyers of those shares must think that sales and profits will soar and OpenAI will join the Mag7 as a listed company with a multi-trillion dollar valuation. That could happen, but it is just as likely that we are close to peak-OpenAI. The list of problems that Sam Altman has to surmount to get to a multi-trillion dollar valuation is very long, and not easily ignored.
Commoditisation: OpenAI is not operating at the level in the value chain where the money will be made. In 1998 Cisco, the maker of internet routers, was the most valuable company in the world at a $550 billion valuation. Investors thought that the immense value created in the roll out of the internet would be captured by the makers of the plumbing – Cisco, Alcatel, Nortel, etc.
But instead, the makers of applications that sat on top of the plumbing – Amazon, Meta, Google, Microsoft, Oracle, Netflix, Salesforce, etc. made the money. The same is going to happen in the AI era. The models will be high capex, commoditised utilities. The applications that sit on top of the models will make the money.
What will those applications be and who is going to make them? – we don’t know. Their creators are currently in dorm rooms somewhere. We do know that OpenAI is at the wrong level in the value chain to make the money.
Revenue model: Another lesson from the internet era is that just because you create value for users of your product doesn’t mean they will pay for it. Users keep looking for the free option. Nobody pays to use Google, Facebook, Wikipedia, Youtube. Most people don’t pay to read newspapers online.
Natural monopoly: OpenAI would love the market for consumer AI to tip in their direction — following the Google search playbook. Google was so superior in search that everyone used it and all the advertisers in that space followed the users. Similar story with Facebook, YT and Ebay. But OpenAI won’t be able to replicate that.
Deep pockets: Alphabet, Microsoft, Nvidia, Amazon, Apple, Meta, all have massive cash flows to sustain them through the immense levels of capex needed to stay in the game. At some stage, excitement about AI will hit a long flat spot and funding for OpenAI will dry up. Then it will be all over for them. They will fall behind, talent will leave and they will be the Yahoo, Netspace, WordPerfect, take-your-pick, of the AI era.
We could go on. I have not even mentioned the problems with their not-for-profit status which prevents them from listing and limits the upside to investors.
Copyright Sam Wylie, 22 September 2025