OpenAI's Wild Ride
Rachel Metz
"Chaos isn't a pit. Chaos is a ladder." Perhaps Littlefinger's famous line in Game of Thrones should be OpenAI's new slogan?1 We seemingly simply can't make it a week without something wild happening within the company. And at the same time, the company keeps pushing forward, on the verge of becoming the most highly valued American startup, still shipping product...
But the past 12 hours have been so chaotic that it's hard to even pick something to link to here. Perhaps it should be the bombshell news that CTO Mira Murati suddenly stepped down from the company? Or the news that the Chief Research Officer did as well – alongside another key researcher? Or the notion that OpenAI is closing in on shifting to a for-profit entity? Tied to that, there's news that Sam Altman may be given an equity stake in the business for the first time.2 But I'm going to use Rachel Metz, Dina Bass, Edward Ludlow, and Shirin Ghaffary's story for Bloomberg as the main link here because they nicely tie everything together and seemingly pinned down the equity stake...
OpenAI is discussing giving Chief Executive Officer Sam Altman a 7% equity stake in the company and restructuring to become a for-profit business, people familiar with the matter said, a major shift that would mark the first time Altman is granted ownership in the artificial intelligence startup.
7% of $150B – the rumored valuation of the most recent funding coming together – is just over $10B. That's probably not a coincidence. [See update below]
The real question is if any of the timing of all of the rest of the above is – in particular, the departure of Murati. While she wasn't technically one of the co-founders of the company – which keeps the current number of original team members still at OpenAI at 2 out of 11 (yes, Greg Brockman is still on leave – does this make it more or less likely that he'll ever come back?) – she was clearly a vital part of the team. So much so that when Altman was infamously ousted by the board, she took over as interim CEO for the few days he was gone. By most accounts, she was not a part of that coup (nor was she on the board that staged it), but instead was viewed as a stabilizing force so the company didn't collapse overnight.
All that said, it is worth noting this excerpt as reported by Deepa Seetharaman, Berber Jin, and Tom Dotan for The Wall Street Journal:
Many at OpenAI expected Murati would leave last year, after she played a significant role in the short-lived ouster of Altman as CEO. Murati had previously approached some of OpenAI’s board members with concerns about Altman’s leadership, according to people close to the company. She described some of his leadership tactics as psychologically abusive and said she was likely to leave, according to people close to the company.
Instead, when Altman was reinstated, she shifted back in the CTO role, where she has remained for the past many months. And beyond Altman, she's been the face of the company in the press. Sometimes to her own detriment. But clearly she's beloved inside the company – back to the Bloomberg report:
On Wednesday, many employees were shocked by the announcement of Murati’s departure. On the company’s internal Slack channel, multiple OpenAI employees responded to the news with a “WTF” emoji, according to a person familiar with the matter.
Clearly, this was not some sort of planned transition – as Altman's statement (which was actually a re-statement to include the people beyond Murai leaving– read into that what you will) highlights, opening with (literal highlights are mine):
Mira has been instrumental to OpenAI’s progress and growth the last 6.5 years; she has been a hugely significant factor in our development from an unknown research lab to an important company.
When Mira informed me this morning that she was leaving, I was saddened but of course support her decision. For the past year, she has been building out a strong bench of leaders that will continue our progress.
And, to his credit, he closes by acknowledging that this isn't "natural":
Leadership changes are a natural part of companies, especially companies that grow so quickly and are so demanding. I obviously won’t pretend it’s natural for this one to be so abrupt, but we are not a normal company, and I think the reasons Mira explained to me (there is never a good time, anything not abrupt would have leaked, and she wanted to do this while OpenAI was in an upswing) make sense. We can both talk about this more tomorrow during all-hands.
He notes (again, in the re-statement) that while Chief Research Officer Bob McGrew and research VP Barret Zoph are also leaving, they all "made these decisions independently of each other and amicably". So again, he's chalking these three departures up to coincidental timing. That's certainly possible, but it's also hard to overlook all of the other recent departures. Someone has to ask: what is going on here? The talking point is clearly that it's normal company growing pains exacerbated by the fact that OpenAI is "not a normal company," in Altman's own words.
But why now?
I think the reasons Mira explained to me (there is never a good time, anything not abrupt would have leaked, and she wanted to do this while OpenAI was in an upswing) make sense.
That seems fair enough, as OpenAI not only just shipped o1 – aka "Strawberry" – out the door, and got the vocal mode for 4o, long delayed, finally live. And presumably the massive fundraise is part of the "upswing" too. Then again all three of these departures were giving interviews to the press in recent weeks about some of these features and the company. Why did Murati pick a seemingly random Wednesday in September to make the call? And why was it the exact same day that Altman's potential equity stake (tied to the shift from a non-profit) leaked? These are strange coincidences, if nothing else!
As I wrote – checks date – not even two weeks ago:
I mean sure, it's true that all startups have growing pains, but it's not "common" for startups to have these types of growing pains. Most don't have boardroom coups and have to deal with the loss of nearly the entire founding team – certainly not this near to achieving breakout success. And most don't have co-founders suing them, twice. And most don't have this building tension with their main backer, at least not this early, and again, in the success state. And most don't have a main backer that is one of the largest companies in the world, which is also competing for business and breakthroughs. And most don't have that main backer hold no actual equity in the business. And most don't have the co-founder and CEO also have no equity in the business! Hell, even the boardroom movements – or non-movements – are dramatic.
One might call this all decidedly "uncommon".
Yes, it's all magnified – success does that – but this is also way more to magnify than your usual startup. Part of that is the sector – AI is seemingly driving every narrative in tech right now from startups to VC to public companies. Part of it is the swirl around Sam Altman, and yes, that history with Elon Musk. But the largest part of it all still seems to stem from the fact that the company was set up one way and is now clearly a very different company.
That's obviously what led to a lot of that early team leaving – with a number either starting competitive projects or otherwise joining competitive projects. And it's what led to the boardroom coup. The merits there still aren't fully clear, but without question, that would not have happened if OpenAI were set up as a "normal" startup. And now as they try to morph into more of a "normal" startup, perhaps in tandem with this new financing, the tensions are all boiling over.
Subsequent reporting will undoubtedly dig up more – or perhaps less, if these are all just coincidental. But the most telling thing may end up being where Murati winds up. She presumably can't and won't go anywhere immediately (Bloomberg notes that she doesn't have a firm exit date yet). But if she ends up at one of the competitors, that probably won't be a coincidence!3
One more thing: Be prepared for Elon Musk to go berserker mode if/when the for-profit shift and equity package for Altman is complete/confirmed.
Chaos isn't a pit. Chaos is a coincidence.
Update September 27, 2024: Sam Altman is now reported to be denying that he's going to receive a "giant equity stake" in OpenAI. Well, that's not exactly true. He's just saying there are no plans for it, right now. But board chair Bret Taylor admitted on the record to CNBC that it's being discussed, but without specific figures or timelines. That's undoubtedly because any such manuever would likely be tied to shifting to a for-profit entity, and that's likely to take a while -- well into next year, per previous reports. Anyway, there's a lot of wiggle room in these statements and denials, of course. But they are worth noting in reference to the very specific "7%" number Bloomberg cited yesterday...
M.G. Siegler
M.G. Siegler
M.G. Siegler
M.G. Siegler
M.G. Siegler
1 To go with the new logo?
2 As I wrote 10 days ago:
One thing that hasn't been mentioned: what this will mean in terms of equity for Altman himself? This was rumored months ago – making sure he has some "skin in the game", and presumably that will be a part of this shift as well.
3 California, of course, doesn't recognize non-competes. But presumably OpenAI would try to incentivize her not to join one of the core competitors. Though it's also not like that stopped others! And, of course, this is just wild speculation. Maybe she'll start her own thing too. Maybe she'll do something completely different. Maybe!