The Failed OpenAI Coup and the Game of Machine Thrones

OpenAI: Sam Altman and the Old Board

Perhaps writing about this story on this blog doesn’t really do anything, but I wanted to make sense of this incredible series of events for myself. I previously wrote about my opinions on AI, seeing it as a solution to corruption and bureaucracy in government and business. But now, with the conclusion of this failed coup d’etat attempt in OpenAI, it looks like it’s no longer about the advancement of technology and humanity first and foremost. Smartphones are now a ubiquitous commodity and cryptocurrency is becoming more and more synonymous with fraud, so artificial intelligence is now the hottest ticket to billions in the 2020s.

As of this writing, this news is so last week, but I still had to write about it because it’s such an interesting story. It’s also a reminder that I shouldn’t sit on my writing because things can change very quickly in this day and age. Despite that, I think it’s always a good time to talk about artificial intelligence as it’s likely the most important technological advancement in the 2020s. It’ll be quite a wonder to see how far it’ll go by the time this decade ends.

But for now, just as 2023 is about to end, we may have just witnessed a sea change in AI research and development. Things will never be the same.

Disclaimer: I don’t know shit about artificial intelligence and corporate business. I’m just a crummy blogger who wants to better understand things, and the main way I do that is by researching and writing about them. If you think I got anything wrong, please tell me about it in the comments section. Reader discretion is advised.

What is OpenAI?

You can google ‘OpenAI’ and get a good description of what it is. Instead of coming up with my own explanation of what OpenAI is, I asked ChatGPT what it is.

OpenAI is an artificial intelligence research laboratory consisting of the for-profit OpenAI LP and its non-profit parent company, OpenAI Inc. The organization was founded with the goal of advancing digital intelligence in a way that benefits humanity as a whole. OpenAI is notable for its commitment to conducting research in the open and sharing its findings with the public.

OpenAI has developed several iterations of large language models, such as GPT-3 (Generative Pre-trained Transformer 3), which is one of the most advanced language models to date. These models are capable of understanding and generating human-like text based on the input they receive.

The organization has both research and safety as key pillars of its mission. OpenAI is actively involved in advancing the field of artificial intelligence, and it also works on ensuring that AI technologies are developed and deployed in a safe and beneficial manner. OpenAI collaborates with other research and policy institutions and seeks to address global challenges associated with artificial intelligence.

ChatGPT is OpenAI’s most well-known creation and the catalyst for the following events. Without its explosive popularity since its public release a year ago (as of this writing), Microsoft wouldn’t have expressed their interest and Sam Altman wouldn’t seem like a lunatic to the OpenAI board and they’d still be hobbyists trying to create a Dota 2 bot that makes The International look like amateur hour.

Summary of Events

The following events happened within a span of five days. This story is worthy of a Netflix documentary, hopefully directed by someone who won’t gamble the production budget on stock options. In the meantime, other companies either announced their own AI products to capitalize on the hype or dissolved their own AI teams to avoid making the same mistakes.

I’m not sure if I got the chronological order of events right, but here’s the gist:

  1. CEO fired by board for being “not consistently candid in his communications.”
  2. President demoted from board and second CEO assigned.
  3. Demoted president quits company.
  4. Employees threaten to quit if the first CEO does not come back.
  5. Board implored the second CEO to bring back the first CEO.
  6. Microsoft CEO assigns fired CEO as head of new AI division.
  7. Board fires the second CEO for not being able to bring back the first CEO.
  8. Board assigns third CEO.
  9. Investors sack old board.
  10. First CEO comes back and the third CEO leaves.
  11. New board has two new members, only one of the old board remains.
  12. Things go back to normal, but not really.

The People Involved

More important than understanding what happened is knowing our cast of characters, the pieces on the chessboard, the inmates of the asylum. No matter how everything seems to be about artificial intelligence in this situation, it’s really about politics and human pettiness. The strongest motivator throughout human history has always been envy and spite.

All of these things center around Sam Altman. The old board wants him gone, and the forces opposing them want him back because they’re invested in the non-profit OpenAI through its limited for-profit company. After the events of this fiasco, any pretense of this whole thing being for the good of humankind is gone. People are in this for the money, same with any other tech.

OpenAI Executives

  • Sam Altman — Chief Executive Officer; cofounded OpenAI in 2015; main spokesperson; AI accelerationist
  • Greg Brockman — President, head engineer; responsible for ChatGPT code
  • Mira Murati — Chief Technology Officer; leads research and product safety; 2nd CEO

The Old Board

  • Ilya Sutskever — Cofounder of OpenAI; chief scientist; AI decelerationist; changed sides in the midst of the crisis; really should accept reality and shave his whole head
  • Adam D’Angelo — Joined board in April 2018; CEO of Quora, whose business was likely negatively impacted by popularity of ChatGPT
  • Helen Toner — Joined board in September 2021; director of strategy and foundational research grants at Georgetown University’s Center for Security and Emerging Technology; AI decelerationist, published a paper criticizing Altman and timeline of OpenAI product releases
  • Tasha McCauley — senior management scientist at RAND Corporation; AI decelerationist; proponent of effective altruism; best known as wife of actor Joseph Gordon-Levitt

Microsoft

Others

  • Emmett Shear — Former CEO of Twitch; 3rd CEO; known to be a blithering idiot
  • Paul Graham — Co-founder of the startup accelerator Y Combinator; Sam Altman’s old boss; hinted at Toner and McCauley (maybe) being dead weight
  • Elon Musk — Needs no introduction; Former investor in OpenAI 

The New Board

  • Bret Taylor — Co-CEO of Salesforce
  • Larry Summers — Former US Secretary of Treasury
  • Adam D’Angelo — The last member of the old board still standing; the “Littlefinger” of OpenAI

There could be more of them, but these are the ones you should really care about.

Why Was Sam Altman Fired?

There are only rumors and speculation on why the old board fired Sam Altman. The following are theories, which may either be true in some way or painfully wrong. We have no way to be sure unless someone from the inside with knowledge of what really happened actually talks about it, and even that would be nothing more than anecdotal at best.

Or better yet, more than one of these theories may be simultaneously true. That’s equally likely.

The weird corporate structure of OpenAI

With the way OpenAI works as a company, being a nonprofit that’s being funded by a limited for-profit wing, Altman’s position as CEO came with a lot of baggage and perhaps even conflicts of interest that the board wasn’t comfortable with. This is a fairly weak reason since the company had been fine with that for six years. How can that all of a sudden be broken?

Altman’s Side Hustles

He has side projects like Worldcoin, which trended on Twitter for a day or two supposedly due to long lines during its launch. He’s also seeking funding for an AI chip making venture to free OpenAI from their dependence on Nvidia. He also tended to act on his own on behalf of the company, following the adage of “It’s better to ask for forgiveness instead of permission.”

In OpenAI’s first developer conference in San Francisco, Altman announced new features for ChatGPT, including allowing users to make task-specific chatbots that may function independently from ChatGPT in the future. This was seen as a safety red flag, which may have spurred Sutskever and the board to step in and deal with the seemingly rogue CEO.

Adam D’Angelo has a Vendetta

As the CEO of the social question-and-answer website Quora, the website that helped kill Yahoo Answers is likely being killed by ChatGPT. The former Facebook executive and now board member of OpenAI did seem like he had a bone to pick with the company and Sam Altman, the poster boy for ChatGPT.

It also didn’t help that D’Angelo was at the helm of Poe, an AI chatbot app that effectively got snuffed by the release and viral popularity of ChatGPT.

However, as the only old board member left standing, it does seem like they still trust him, despite the circumstances that led to the crisis in the first place. Why would they keep him on if he’s so bad? What is he bringing to the table that the other two ex-board members didn’t?

Helena Toner hates Sam Altman

When the crisis started, Toner immediately clammed up, set all of her social media accounts to private, and lawyered the fuck up. As mentioned, she published a paper on CSET that criticized OpenAI’s direction under Altman, which was a pretty ballsy move.

According to multiple sources, Altman wanted to kick Toner out of the board for quite some time. If there’s someone who had even more of a reason than D’Angelo to get rid of Altman, it was Helen Toner.

AGI breakthrough with safety concerns being rushed by Altman

This seems to be the most likely reason. The speculation is that Sutskever saw something that spooked him so much that he told the rest of the board about it. That gave everyone else a casus belli to push for the firing of Altman. To support this claim, various figures including Elon Musk have said that Sutskever possesses a strong moral compass and is not one to play political games and seek power.

Is Everything Really Back to Normal?

Not really. Whatever pretense left of OpenAI being a nonprofit organization is long gone. Everything that OpenAI works on now is for the money, so whatever science fiction idealism that may have driven the old board (or used as a convenient excuse) to fire Sam Altman is now purged from the company. It’s the only way AI will truly become the next big thing in tech.

You can’t hire the best with only ideals and whatever the hell “effective altruism” is. You don’t study computer science and engineering just to serve humankind. They’re going there to get rich.

With Microsoft completely behind Sam Altman and his team, all this coup really accomplished is make the old board look silly. And yet, Adam D’Angelo is still a board member. He must have done something right to survive the fallout somehow. He survived while his colleagues in the old board didn’t. There’s a story there we should find more about in the future.

But for now, everything is back to semi-normal. Ilya Sutskever is no longer on the board, but is still an employee. Adam D’Angelo is the only old board member who is still on the board. Helen Toner and Tasha McCauley are completely out of OpenAI.

As one of the top comments in the Slidebean video says, “(The) Board thought Sam was a bug, employees affirmed he’s a feature.”

I leave you with this image I found on this Reddit thread of Ilya Sutskever as the harbinger of AGI.

Ilya Sutskever as the Harbinger of AGI

Got Feedback?

Have something to say? Do you agree or am I off-base? Did I miss a crucial detail or get something wrong? Please leave whatever reactions, questions, or suggestions you may have in the comment section below.

You may also like/follow and leave a message on either Facebook or Twitter. Please subscribe to both the Avoider.net YouTube channel and my personal YouTube channel, as well as my Twitch channel for more content. Thank you for dropping by.