The $60,000 "First Customer Ship" Clock

 After Deliberative Analysis: the Ouster of Sam Altman from OpenAI

Yesterday, at noon Friday, 17 November 2023, one of the most surprising announcements in the tech world was posted on the blog pages of OpenAI. The announcement's headline closed with, "Sam Altman departs the company".  In practical terms his ouster wasn't too much different than a Twitter or Square employee. He turned up to a video meeting and "was immediately fired" according to board chair and OpenAI founder and president Greg Brockman.

Those words fell like a meteorite on the tech world.  Reuters had a four-reporter story up before 10p. The New York times followed immediately. The Atlantic had a story up by the morning . YouTube AI reporter Matt Wolfe was up even sooner with the news.  By the morning in California, France's digital minister opened the door to ex-patriatism. To say this was a stunning announcement is an understatement.  I'm not going to recap who Sam Altman is or why his leaving the hottest AI company on the planet and in history in the year AI has made the biggest actual impact on the workings of society is significant.  But I like the line in the Atlantic article, 

"It’s hard to think of an analogue to Altman’s firing; it’d be as if Bill Gates was abruptly shown the door at Microsoft in 1995"

Altman's reach goes far beyond Silicon Valley; he's a tech rock star in the same vein--if not mold--of Elon Musk, as far as name and perhaps face recognition goes. To the mold, I'd have him more akin to a casting of Steve Jobs, with a deep sensitivity to impact of the tasks he set for OpenAI and their impact on the world. To be fair, Musk's and Altman's visions play with the same goals: the transformation of our civilization.   And techies and AI followers since befor 2023's blowout year will know that Musk was an OpenAI founder--they're not just abstract same goals, they're in the same sandbox.  But Altman has never been as belicose as Musk. He is soft-spoken. Even at the congressional hearings (urged by Altman--what CEO asks a congressional comittee to get all up in their business) he often took a pleading tone, conveying urgency in front of his always sympathetic countenance. 

So why the very sudden (to the public's knowledge) fall from grace? Let's start at the top and look at what OpenAI said in it's announcement. It's common for a board Chairman to step down after a CEO is fired, especially if the were co-founders. It's a clear sign of loyalty and an equally clear signal that Brockman didn't vote with the majority for the ouster. Brockman was interestingly, not part of the board meeting at which a vote to fire Altman was passed. He was told minutes later, "about the same time as the blog post" (NY Times). 

The Board Votes -- employees of OpenAI versus a majority of "independents"

So there were at most four members of the board at the meeting. OpenAI chief scientist Ilya Sutskever, independent directors Quora CEO Adam D’Angelo, technology entrepreneur Tasha McCauley, and Georgetown Center for Security and Emerging Technology’s Helen Toner. If we make the (admittedly huge) assumption that Sutskever did not vote 'yea', that left 3 against 2. We also might assume the board has rules of quorum and could assume the "2" would include Brookman.

This is the most obvious case. Then news emerged that this was a "coup". Implying someone within the company was responsible for outing the leader.  That could only have been Illya Sutskever. Wired Magazine reported this today using the exact words, "board coup".  Ars Technica supports this view, that Sutskever aligned with the independents. 

So all three independents and perhaps Sutskever--and interviews with Sutskever will be so very interesting to read--would have had to been aligned to take such a dramatic step.  What did they see? What evidence was under review? 

Information must have come to light that could not be ignored by the board members who, by the text of the announcement, hold "the fundamental governance responsibility of the board to advance OpenAI’s mission and preserve the principles of its Charter."  And what is this charter? It can be read in its entirety, but it is very telling that the announcement excerpts a portion of it as they announce they are ousting their--by many accounts--most important member of the company. They wrote, (interestingly self-quoting in that weird "we're the PR-arm, they're the board, they didn't say any of this directly" CYA mode),  the board of directors said:

"OpenAI was deliberately structured to advance our mission: to ensure that artificial general intelligence benefits all humanity. The board remains fully committed to serving this mission."

What could have violated this tenent, which must include the requirement for "safety" in AI development ("safety" being a very serious technical term to avoid the robot overlord scenario as well as lesser harms to individuals and society).  

Perhaps it was products. Rushed--in someone's view--to market? Revealed on OpenAI's Dev Day not long ago? It definitely wasn't malfeasance, or other potentially criminal causes according to a statement made by the COO Brad Lightcap. 

But direct to consumer or even vertical enterprise software (we'll help you build your own models, touts OpenAI) is chump change for a leading AI company. Altman had already scored billions-with-a-'B' dollars from Microsoft. "Single-handedly." What was the next move?  Apparently taking the company public or something resembling a SPAC.

Where was Sam? Was it The Money?

The firing was widely reported to have made Microsoft CEO  Pachai "furious", though he later issued a statement fawning "full support" for interim CEO (and perhaps Sutskever ally?) Mira Murati and team.  Pachai's folk probably whispered to him that Altman's deal for $80b in owner share conversion was probably not happening soon after Altman losing any say-so. Something that might have impact on the power that Pachai's team weilds as bare-minority stakeholders (49% of the public entity portion of the OpenAI organization stack). 

OpenAI's holding structure, ©️OpenAI

Because that's what Sam has been spending a not insignificant amount of time doing. Some of the news reporting on his firing focused on him being "OpenAI's best fundraiser", which, to be fair is the most important job for a CEO of a privately held company. Especially one in a market where the main 4 competitors have a combined valuation larger than Japan--the country's--GDP.

One wonders how much money can be attracted to a company which has--in the eyes of capitalists--hamstrung itself with ethics in its charter? Did the OpenAI board get a close look at documents which had not been shared with them by Sam on how the reins were going to be lifted in order to keep their lead against Meta, Alphabet and one presumes, eventually, Musk's Anthropic, or Tesla and/or Apple? Was it going to be a raise of the profit cap on OpenAI Global, LLC which would have demanded a compressed development and hard date release schedules to align with the more "normally run" (i.e., shareholder feduciary) Fortune companies?

There is already buzz that Altman is going to immediately pivot into a project he was pursuing at the time of his ouster. Burying the second part of this story, Greg Brockman, OpenAI President and Chairman quit along with a few other senior folk including engineers. So keeping some momentum is not improbable for Altman and crew.  Stories this afternoon suggest it would be something he had made public interest in earlier this year: hardware for AI devices. Some say an AI chip to compete with nVidia offerings and Tesla's Dojo design, but a reported partnership talk with Jony Ive, former design chief at Apple and a Silicon Valley rock star in his own right, suggests something user facing.  Maybe an AI appliance to join the desk, home, independent of a GPT-4 window or app on your laptop or pocket computer?

This, being written a scant day after the events that transpired in San Francisco, is an evolving story. Hearing from those who remain at OpenAI is going to flesh out the details until everyone zips up for competitive reasons. 

Despite the bum's rush, Altman's own star might still be on the rise. Everyone loves a good recovery story.  Altman is already a serial entrepreneur with other AI fire irons already being bellowed. And lest we forget, Steve Jobs returned to Apple after it stumbled, nearly to extinction, and brought it back to glory.