OpenAI 2.0 - Altman Returns
Drama concludes, board is fired, OpenAI re-unites and moves on
The Fate of OpenAI - OpenAI 2.0
I had been wanting to write this week about GPTs, but the drama at OpenAI has been a distraction and kept me wondering about the fate of GPTs. Well, after a number of enough plot twists to make a full Netflix season out of this past weeks events, Sam Altman is back. Altman 1, Open AI Board that fired Sam 0.
The twists:
Friday - the firing: CEO Altman fired, board announces ‘transition’; President and co-founder Brockman quits; Ilya has employee meeting, questions about ‘coup’ arise. Mura Murati named interim CEO.
The Weekend - the split: Employee revolt brews; tweet campaign “OpenAI is nothing without its people.” Altman negotiates return; talk of Altman setting up another company. Open letter from employees to board demands the board resign, and remarkably, Ilya Sutskevar, co-founder and board member, is on the signatory list. Altman returns to OpenAI with a guest badge on Sunday to negotiate a return, but no deal is made. Microsoft announces hiring of Altman and Brockman, and invites anyone from OpenAI to join. OpenAI Board hires former twitch CEO Emmett Shear as new OpenAI CEO. It looks like it’s Altman is gone for good.
Monday - Tuesday - turmoil: On Monday, Microsoft went UP 2% on news of the coup of getting the entire OpenAI crew in-house. The board continued to be opaque about why Altman was fired. There was miscommunication or distrust, but the board fails to be clear and convincing about what it was. Various stories were floated, finally centering on Helen Toner and a disagreement with Altman over an academic paper that praised Anthropic over OpenAI. With over 700 of 770 employees calling for the return of Altman and the firing of the board. This situation for the board became untenable, as it looked like firing their CEO would mean the destruction of OpenAI as a company.
Wednesday - resolution: A deal is struck. OpenAI board relents. Sam Altman gets reinstated as CEO — The board that ousted him gets fired and replaced. A new board at OpenAI replaces the old board, and consists of Larry Summers, Bret Taylor, and Adam D’Angelo (the Quora CEO who was on the board previously).
Sam put it all on X, and he apparently remains allergic to capitalization:
i love openai, and everything i’ve done over the past few days has been in service of keeping this team and its mission together. when i decided to join msft on sun evening, it was clear that was the best path for me and the team. with the new board and w satya’s support, i’m looking forward to returning to openai, and building on our strong partnership with msft.
TechCrunch takeaway is that “OpenAI will benefit from unity of purpose with Sam Altman’s return.” The core problem of a non-profit at the epicenter of a technology revolution that will be worth trillions is not fully resolved, but this exposed that a small and weak board was making decisions and was not answerable to anyone.
They will make the board better, they will keep OpenAI intact, and they will get back to shipping product and developing further.
This weak board exposed a problem to the VCs and Microsoft who put billions into the profit-making side of OpenAI. They need better governance and better assurances on their investments and partnerships. This outcome surely was better than the near death-sentence the decimation and division of OpenAI’s people would have been.
Satya Nadella almost had his own AI team, but now has an OpenAI more committed to working with Microsoft than ever. He comes out a winner, as he stood by the OpenAI CEO when they needed him, forcing an untenable situation for the board that forced a return of Altman. His praise of the new OpenAI board indicates he is happy with this resolution:
We are encouraged by the changes to the OpenAI board. We believe this is a first essential step on a path to more stable, well-informed, and effective governance. Sam, Greg, and I have talked and agreed they have a key role to play along with the OAI leadership team in ensuring OAI continues to thrive and build on its mission. We look forward to building on our strong partnership and delivering the value of this next generation of AI to our customers and partners.
So, we are back to a restored OpenAI. OpenAI 2.0 will be stronger and more unified, and clearly the support for Altman by OpenAI’s team means that his vision will prevail in what is clearly his company.
The Fate of AI
What about us? What about AI? Will OpenAI 2.0 be a better OpenAI? Who governs the future of AI and the path to AGI?
Claims of “Effective Altruism” and the “good of humanity” we have seen get used by con-men like Sam Bankman Fried and as a cover for self-serving selfishness. Google said “Don’t be evil” yet one can find them as guilty as tobacco companies of profit-making and manipulation over people. There would be more sincerity in those who admit they are acting in self-interest and don’t really know what’s best for the world.
Perhaps a divided OpenAI might have led to more competition and openness that would have been better for the world. We won’t run that experiment after all. On the other hand, had Microsoft fully owned the OpenAI team, the best AI models in the world would be buried in the most valuable companies in the world. That might have been too much concentration in power.
As it is, just this week, InflectionAI has a new model, Inflection 2.0, that they claim is close to GPT-4. Anthropic announced Claude 2.1, with tool use, system prompts and a 200k context length with improved performance across the context window. The rest of world isn’t standing still.
LLMs are to the AI age what the browser was to the Internet age. It is the substrate and foundation for a lot of the value creation in the AI space, but much more will be built on top of foundation models to make AI truly amazing.
The OpenAI drama is over, now back to our regularly-scheduled AI Revolution.