What a genius!
Absolutely brilliant.
Only it’s not a person, it’s OpenAI’s latest frontier AI model – o3.
Source: Tracking AI
Shown on the far right is OpenAI’s “o3” model, released just a few days ago, along with its updated 04-mini model.
o3 is now OpenAI’s most advanced reasoning artificial intelligence model. And as we can see above, it is scoring at a genius level on the Mensa Norway IQ test, unseating Google’s Gemini 2.5 Pro Experimental model.
But of course, that’s just one benchmark used to evaluate AI models for intelligence.
The race towards artificial general intelligence (AGI) has become so intense since xAI stunned the industry with its release of Grok 3. Grok 3 has lit a fire under an already frenzied sprint to be first to AGI.
Every week, we’re now seeing major releases from at least one major player of AI models with remarkable performance. It’s really a neck-and-neck race right now with each of the major models outperforming the others on one or two common AI performance benchmarks.
Source: Artificial Analysis
There is no clear “winner” right now, but what is clear is that OpenAI, Google, and xAI are jockeying for first place, and Meta’s Llama and Anthropic’s Claude have fallen a bit behind.
What a ridiculously exciting competition to watch. It’s like the computer science version of the NHL playoffs, and we’re in the semifinals. But this competition will happen only once in history. There will only ever be one company to first achieve AGI.
But many will follow after that…
As I’ve written many times before, this isn’t just a massive research and development effort. It’s not an experiment. It’s not a speculative bubble, though I know many wrongly believe it is.
This is already a massive business with a clear path towards high gross margins and an avalanche of free cash flow.
And this is precisely why the levels of investment in AI continue to materially increase with each passing month. Unlike speculative bubbles, AI already has the hard numbers to back it up.
No better example is OpenAI, a company that we know well at The Bleeding Edge, as we’ve been tracking its progress from the very beginning.
Last year, OpenAI generated an incredible $3.7 billion in revenue, and this year was expected to see $11.7 billion, a crazy number representing more than a tripling in revenue.
Source: The Information
But those are old numbers. The projection as of fall last year was $11.7 billion. But as of a couple of months ago, OpenAI is now projecting $13 billion in revenue this year – an 11% increase from the previous projection.
And while the majority of that revenue increase is coming from ChatGPT-related revenues, we can see on the right in the chart above that revenues related to agentic AI services are expected to become a multibillion-dollar business this year.
What an incredible insight this provides. It tells us that we’re on the cusp of a wave of agentic AI services that will become available to anyone with an internet connection. And we should expect that the majority of the utilization, and thus revenue, will come from the U.S. market this year.
That’s due to the combination of the U.S.-centric development of agentic AI services, as well as the regulatory environment for artificial intelligence.
The current U.S. administration is very supportive of the development and commercialization of AI, whereas many other markets around the world, like the EU and China, have very strict regulations in place that will interfere with the adoption of agentic AI services.
As for the overall direction of OpenAI, the charts shown earlier are obviously up and to the right. OpenAI’s revenue forecasts for 2029 have popped from $100 billion to $125 billion, a large 25% increase in forecast just since last fall.
Better yet, gross margins are expected to reach nearly 70% by 2029 from where they are today, just under 50%. OpenAI estimates that it will generate $12 billion in cash that year.
This is precisely why OpenAI is currently valued at $300 billion in its latest venture round this March. OpenAI raised $40 billion at a $260 billion pre-money valuation. The valuation seemed like a lot at that time, but now that we know that OpenAI expects $29 billion in annual revenue in 2026. That’s an enterprise value to 2026 sales multiple of 9 – not unreasonable at all for a company growing this quickly.
What does this all mean? Well, it means that OpenAI will become a trillion-dollar company, probably before the end of 2027… a distinction only 10 companies in the world can claim, and fewer still in so short a time.
Imagine that. Going from its first capital raise in 2015 to a trillion-dollar company by 2027. Amazon Web Services, Infosys, Y Combinator, Peter Thiel, and Reid Hoffman all participated in that initial round.
The valuation of the company wasn’t made public for that investment, but let’s assume it was around $3 billion. Those investors are now up 100x on their original investment, less any dilution.
How can normal investors gain exposure to OpenAI? One way would be to acquire shares in Amazon (AMZN) or Microsoft (MSFT). Amazon has exposure through its ownership via its Amazon Web Services’ investment in 2015. Microsoft stepped up and invested $1 billion in 2019 and an additional $10 billion in 2023.
The $1 billion investment from Microsoft was likely made around a $4 billion valuation, and the $10 billion investment around a $25 billion valuation. That would imply Microsoft is up 75X and 12X, respectively, less any dilution. I seriously doubt that value is reflected in Microsoft’s share price.
Many believed that Microsoft was stupid for investing so much in OpenAI at such inflated valuations. I felt very differently. OpenAI was growing at such a rapid rate with a concrete business model and real revenues coming in that it was clear the valuation would continue to climb.
And at a trillion valuation for OpenAI, Microsoft could potentially receive more than $100 billion in a payout on an exit. That’s a huge sum even for a company worth $2.9 trillion. It’s a lot more than the $71 billion in cash Microsoft currently has on hand and enough to eliminate its debt entirely… Not that Microsoft needs any more money.
If all of this isn’t enough to be bullish about OpenAI’s prospects, the latest antitrust issues with Alphabet’s Google division are presenting what could be an incredible opportunity for OpenAI. A U.S. District Court judge has already found that Google holds an illegal monopoly over the internet search market.
This has resulted in the U.S. Justice Department asking that Google divest its Chrome browser. And not surprisingly, OpenAI has already publicly stated that it is an interested buyer.
What would that buy OpenAI?
Distribution, plain and simple.
Chrome commands about a 56% market share in the U.S. for internet browsers and an incredible 67% market share worldwide.
If OpenAI were able to own the Chrome browser division of Google, it would greatly enhance its ability to drive even more revenue to its business and potentially “acquire” a majority of all generative AI and agentic AI revenues globally.
Who knows what OpenAI would be worth then?
Jeff
The Bleeding Edge is the only free newsletter that delivers daily insights and information from the high-tech world as well as topics and trends relevant to investments.
The Bleeding Edge is the only free newsletter that delivers daily insights and information from the high-tech world as well as topics and trends relevant to investments.
