The piece I did the other day on SpaceX is already one of the most popular I’ve done since leaving Investorplace in late 2024.
But it’s part of a pattern that has been building for decades now, of oligarchs taking more-and-more cash from investors, while giving less-and-less equity and control back in return.
It’s a part of tech’s history that founders need control of their companies to speed decisions down the chain. Having one decisionmaker, and one vision, means decisions are made faster and the vision is sustained.
This can be healthy, but it can become toxic. As the numbers required to play the Great Game have grown, and the speed with which companies must move has decreased, CEO power has jumped, not just at the expense of investors but investment bankers.
Founders’ shares, dual-share structures, and trading just a tiny sliver of equity in a giant company are all used to leverage the CEO’s power. These games were pioneered by the men who are now centi-billionaires. But the abuse has spread down to otherwise mediocre men and women who can do a few deals, create billions in market cap, then “retire” to their mega-mansions. Successors must clean it up, aided by consultants hollowing out the companies to keep the stock price up.
I mention this only because it’s the scams done by lesser men that justify those of the big game players. The problem is, as in the case of SpaceX, that the trick is becoming obvious, even to a mere reporter.
Which brings me to OpenAI.
Copycats Fail
There are pieces of SpaceX, like Starlink, the rockets, and Tesla, that are profitable. There’s nothing profitable at OpenAI.
Om Malik has a name for the game Sam Altman is playing. He calls it the announcement economy. Manipulating the message, turning corporate news into a magic show, can now make billions fall into even a failing start-up.
To make the game work, growth must matter more than profit. The trick is to make investors focus on the first while ignoring the second. This is what the Great Game is all about. To dream is to dare.
This may be where Altman, and the trick, fails.
While OpenAI has paying customers, and $25 billion in annualized revenue, it has no profit. Nor does it have an assurance of profit down the road. AI itself is as popular as MAGA right now. The OpenAI IPO is starting to look like tech’s Iran War.
What’s funny is how the Administration has its fingers deep into this cookie jar. Calling Anthropic an enemy has made it Zelensky in the eyes of users, and OpenAI into the Great Game’s Putin. That’s unfair, but in the announcement economy it’s the story that sells, not what lies behind the curtain.
Hiring Peter Steinberger doesn’t give OpenAI OpenClaw. Expect him to be shoved onto the stage at the next OpenAI presentation, because OpenClaw today is what OpenAI was just three years ago. The problem is that OpenClaw really is open source. Steinberger doesn’t own it. OpenAI doesn’t own it. That means Nvidia can easily gain control of its enterprise version with NemoClaw, which adds existing parts of its Agent Toolkit to deliver a version with privacy and security.
Nvidia has already delivered what Steinberger may just be working on.
The Deflating Balloon
I predicted back in October that the AI bubble will bust when OpenAI tries to go public. Unlike SpaceX, which can push Wall Street around because it doesn’t need the money, OpenAI badly needs the money. The deal investors are offered may look much better than Elon’s.
It won’t be.
Look closely at how many other tech giants react to the coming prospectus and road show. Is Nvidia going to throw cash at it? Is Musk? Will Microsoft?
That’s what they’ve been doing in the private offerings, leveraging billions into hundreds of billions by supporting bigger rounds at higher valuations.
Is that going to work again, in a publicly traded issue? Can small investors be convinced this is going “to the Moon” again, based entirely on promises from a man they have been told for years now no one trusts?
I doubt it. But I’ve been wrong before. I am, after all, usually early to the party, and the first to leave the bar.
