Is Microsoft about to get the deal of the century? Or is Sam Altman unloading OpenAI at just the right time?
I don’t usually write about business deals, much less about rumors about business deals, but this one has me scratching my head, and and is actually super relevant to how people on the inside - both at Microsoft and OpenAI are viewing the future of AI.
According to a rumor published at the website Semafor, and then repeated by Reuters, Microsoft is preparing to put 10 billion dollars (possibly partly in the form of cloud credits?) into OpenAI, valuing OpenAI at $29 billion dollars.
Of course, whether you think $29 billion is a sensible valuation for OpenAI depends a lot of what you think of their future. (Revenue so far is rumored to be more like $29 million/year, so you have to think things will soon get a lot better to make this sort of deal). As in a lot of tech deals, there is a lot of uncertainty (eg can ChatGPT actually displace Google? I have argued no, but nobody knows.).
Clearly, there are more zeroes here than in most AI deals. (Google, before they were Apple, bought 100% of Deepmind for something like $600 milllion in 2014). Looked at that way, Microsoft sure is putting in a lot of money.
But things start to get weirder, a lot weirder, if Semafor’s reporting is correct.
To begin with, this isn’t some ordinary business deal in which some investor gets say 20% of a company in exchange for a big bolus of cash and a board seat. Instead, if the reporting is correct:
Microsoft ”would get 75% of OpenAI’s profits until it recoups its investment”. That’s a lot, and not the sort of thing that a rocketship startup would typically agree to.
After that point, when Microsoft makes its money, Microsoft would own “a 49% stake, other investors taking another 49% and OpenAI’s nonprofit parent getting 2%”.
That raises so many questions. One is about OpenAI’s whole history of originally being a nonprofit that changed into a hybrid non-profit/for-profit that seems to be on the verge of for-profit engine for Microsoft—while still getting some kind of protection as an nonprofit. Sharper minds than mine might want to look into that.
Another is about whether this is really even best understood as an investment (in which the original Founders keep some measure of control)l or as as a sale. Is this tantamount to the founders getting out entirely? Do they have retain any control at all? (Are the rumors even correct?)
But, from my perspective, the most interesting question pertains to how OpenAI actually sees itself, when the cameras are off.
On one the hand, being valued at $29 billion dollars is really a lot for an AI company, historically speaking, on the other Altman often publicly hints that the company is close to AGI, which ought to be worth a ton of money if it were true. Just a few months ago the leading investor Elad Gil interviewed OpenAI’s CEO Sam Altman. Gil asks hims how far we are from AGI (artificial general intelligence), and Altman answers without answering, asking a question of his own: how different is current AI from AGI anyway? Altman seems to be implying that we weren’t really that far, in any obvious way.
That’s consistent with his bullish tweet back in April, an hour or so after DALL-E 2 was announced, when Altman posted on Twitter “AGI’s gonna be wild”.
On that theory—on the bullish theory that OpenAI is close to AGI—Microsoft is getting the deal of the century. If it all works out, Altman risks looking like the fool that sold the Louisiana Purchase for a measly $15 million.
But there’s another side to Altman, too, the part that was recently forced to acknowledge that ChatGPT, for all its fame, really isn’t anything like AGI yet:
As I explained to Ezra Klein, and have discussed before in this Substack I personally think the latter, less rosy picture, is closer to the mark. Maybe Altman does, too.
But wait, there’s just one more thing. According to another rumor, reported last week by the Wall Street Journals, OpenAI’s founders are busy preparing to get rid of a bunch of their stake (to the tune of $300 million) in a separate transaction with Founders Fund and Thrive Capital. Often, when insiders sell, it’s taken as a bad signal.
At this point, I just have ask a question: If Altman genuinely thinks OpenAI is close to AGI, why is the company preparing to get out and take so much money off the table, when in principle they might be on the verge of owning the world?
It doesn’t quite add up.
How much would AGI actually be worth? A few years back, PwC estimated that the overall AI market might be worth over $15 Trillion/year by the year 2030; McKinsey published a similar study, coming at at $13 trillion/year. I cited a much lower number on Ezra Klein, about $1 trillion, which I got from an email in 2014 from Peter Norvig; (Stuart Russell once gave me a similar number.)
There is a huge gap between $15 trillion and $1 trillion/a year, about the difference between the economies of China and Indonesia.
But on either story, if you really were close to being first to AGI, wouldn’t you want to stick around and take a big slice of that, with as much control as possible?
My best guess? Altman doesn’t really know how to make OpenAI into the juggernaut that everybody else seems to think he’s got. ChatGPT is super expensive to operate (rumors are about $3 million/day); their revenue so far is modest (perhaps low tens of millionsay too unreliable to being do that anytime soon.
Meanwhile, large language models themselves are quickly becoming commodities; it’s not clear that there is any unique, protectable intellectual property here. What can OpenAI do that Google couldn’t quickly replicate, and perhaps even price cheaper?
Maybe Altman is getting out while the getting is good.
Update: Turns out Semafor was wrong about the deal terms. If things get really really good OpenAI gets back control; I am told by someone who has actually read them, “Once $92 billion in profit plus $13 billion in initial investment are repaid [to Microsoft and once the other venture investors earn $150 billion, all of the equity reverts back to OpenAI.” In that light, Altman’s play seems more like a hedge than a firesale; some cash now, a lot later if they are hugely successful.
Gary Marcus, scientist, bestselling author, and entrepreneur, is a skeptic about current AI but genuinely wants to see the best AI possible for the world—and still holds a tiny bit of optimism. Sign up to his Substack (free!), and listen to him on Ezra Klein.
Thank you, Gary, for your efforts to insert some form of sense and reasoning into a highly charged emotional subject. While LLMs, ML, DL, ANNs etc. are all valuable, none of these individually or integrated provides any basis for tech reasoning. Reasoning is foundational to any true AGI (if that is even possible) and while LLMs are, sometimes, doing amazing things with language, LLMs alone have no understanding of the meaning in what it writes, which requires reasoning.
Wow. $10 billion for an essentially open source technology that pretends to be intelligent but is really a super expensive statistical engine on steroids? Wonders never cease. This is worse than folly. It has fraud written all over it in my opinion. Sorry.
Thank you for another insightful analysis.