OpenAI is the research organization behind ChatGPT, the AI-generated chatbot that took the internet by storm last year for its capacity to have really weird conversations with tech journalists. It’s at the center of Microsoft’s big bet on generative AI tools transforming the world, gaming, and more, and it’s now at risk of imploding after its CEO, Sam Altman, was mysteriously ousted by the OpenAI board of directors and Twitch co-founder Emmett Shear was desperately recruited to replace him. Here’s all you really need to know about OpenAI to appreciate what a clusterfuck the last few days have been.
The research organization has a non-profit wing and a for-profit subsidiary. Sam Altman, as head of the entire organization, raised billions in funding from Microsoft and recently started rolling out experimental AI tools called GPTs. There is the main one, ChatGPT, that will write work emails for you, attempt to summarize articles and research, and often make stuff up.
But OpenAI recently announced a GPT platform where you can create mini, custom GPTs for whatever types of work you need to do. Microsoft just rolled out its own version of this called Copilot early this month, which adds ChatGPT functionality to stuff like PowerPoint. No one knows how any of this will translate into big profits—generative AI is really expensive—but nobody wants to be the last one to find out. The splash made by ChatGPT last year fueled a gold rush in AI speculation, including in the video game industry, and the current chaos underlines the tension between oversight of new technology and Silicon Valley’s penchant for “moving fast and breaking things.”
In the heat of the AI-arms race, Altman was suddenly fired from OpenAI on November 17. Its board of directors announced that “he was not consistently candid in his communications with the board, hindering its ability to exercise its responsibilities.” Co-founder and president Greg Brockman was also removed and said fellow co-founder, chief scientist, and board member Ilya Sutskever was the one who told them they were gone. Microsoft CEO Satya Nadella put out a damage-control tweet saying the long-term partnership between the two companies was fine.
Read More: What People Get Wrong When They Think About Video Game AI
Then, over the weekend, the board started quickly trying to backpedal after staff revolted. Three senior researchers quit, including one that worked on the Dota 2 bot that wrecked pros back in 2018. CTO Mira Murati became interim CEO for roughly two days. There were rumblings of employees threatening mass resignations. The Verge reported on November 18 that Altman was in talks to come back, the board was prepared to resign to make that happen, but then a deadline to come to an agreement passed and everything continued to fall apart.
Today, November 20, OpenAI announced that Twitch co-founder and former CEO Emmett Shear would become its new boss and Microsoft announced it would hire Altman, Brockman, and others to lead a “new advanced AI research team.” Altman would be CEO of the department, equivalent to the title Phil Spencer recently received to oversee all of Microsoft’s gaming business. Wired reported that over 500 OpenAI employees promised to quit unless the board resigns.
“I deeply regret my participation in the board’s actions,” Sutskever, the apparent Brutus in this whole thing, tweeted shortly before the petition was circulated. “I never intended to harm OpenAI. I love everything we’ve built together and I will do everything I can to reunite the company.”
It’s still not clear why all of this happened. Was the board of directors just power hungry? Were there actual principled disputes over the future of AI research and the business-first direction Altman was taking the company? That story will potentially shake out in the days and weeks to come, but for now the OpenAI coup is simply a testament to how rapid and messy generative AI’s ramp up and rollout has been over the past year.
You don’t have to look any further than Shear’s involvement to see that. The former Twitch boss exited the Amazon-owned streaming platform for video game culture and content creator drama earlier this year amid layoffs and malaise. He said he was leaving to focus on his newborn son. Now he’s found a new controversial whirlpool to dive into. “Spending time with him has been every bit as rewarding as I thought it would be, and I was happily avoiding full time employment,” he tweeted today. “I took this job because I believe that OpenAI is one of the most important companies currently in existence.”
Shear said he plans to launch a full investigation into Altman’s ousting and make its findings public. He also claimed the board’s disagreement with the previous CEO wasn’t over safety concerns (everything from AI-generated deep fake porn to weapons of mass destruction) and that there are still plans to continue selling OpenAI’s tools and research moving forward. “I’m not crazy enough to take this job without board support for commercializing our awesome models,” he wrote. This has been far from reassuring to people who feel like the tech sector’s whole AI pivot has thrown caution to the wind.
It’s hard not to wonder if this is the beginning of the end of OpenAI, or an even fiercer and weirder rivalry between it and Microsoft moving forward as both try to chart the future of, and profit from, generative AI’s impact on the internet and the world. One thing is clear, there was little to no vetting of Shear prior to his hiring at OpenAI. Just days prior to getting tapped, he tweeted about how CEOs are overrated. “Most of the CEO job (and the majority of most executive jobs) are very automatable,” he wrote on November 16. “There are of course the occasional key decisions you can’t replace.” That was by far one of the more normal things he’s said this year.
Update 11/20/2023 4:34 p.m. ET: Sure enough, the OpenAI succession story has only continued to get weirder. 404 Media unearthed a massive Harry Potter fanfic that incorporated Shear into its story, while Forbes dug through his thoughts about sex and using AI to write pick-up lines.
But the biggest update of all is that Shear might not even end up running OpenAI at all. The Verge now reports that Altman’s move to Microsoft isn’t finalized, and that the former CEO is still actively lobbying for his old job back and for the OpenAI board of directors to effectively turn control over the non-profit research organization over to him. Even OpenAI’s oldest investor is apparently calling for Shear to make way. As Matt Levine points out, it feels like no matter who ultimately ends up running OpenAI after all this, the money people will be the ones in charge going forward.
Update 11/22/2023 10:42 a.m. ET: Following one of the most dramatic corporate coup attempts in tech history, Altman is now back at OpenAI as CEO and the board members who fired him are gone. Both sides, however, have agreed to pay an outside lawfirm to investigate what the hell actually happened and why. History is written by the winners though, which appears to be Altman, Microsoft, and the new board members.
One big thing to keep in mind throughout this soon-to-be made-for-Hulu docu-series, however, is that no one yet knows if OpenAI will ever be profitable, useful, or even legal. The entire enterprise still depends on paying Microsoft massive amounts of money to run compuations that train ChatGPT on data which a court might one day decide OpenAI doesn’t have the rights to. In the meantime, here is the former head of Twitch, and now also the former head of OpenAI, tweeting about how “pleased” he was to be part of the OpenAI circus these last 72 hours:
