Sam’s Town
Reading Time: 7 minutesWhat the Firing and Rehiring of Sam Altman Actually Means, The deposed OpenAI CEO is back. What did it all mean? Did it mean anything?, Sam Altman was rehired as OpenAI’s CEO. Here’s what it all really meant.
Folks, if you predicted on Friday that the closely watched OpenAI power struggle would end in the most pointless-seeming way possible … well, just look. Late Tuesday night, four days after CEO Sam Altman’s shocking ouster from the A.I. company, we found ourselves (mostly) back where we started: Altman is returning to OpenAI as its CEO, albeit not to its board of directors; Greg Brockman is once again president of OpenAI, but also will not be a member of the board; Mira Murati, who briefly took the helm as interim CEO, is just regular ol’ CTO again; the three researchers who’d stepped down Friday in solidarity with Altman and Brockman are either back at the company or requesting to return; Altman & co. will once again operate with the backing of Microsoft, not as direct employees of the Big Tech pioneer. When it comes to the Main Characters of this saga and their loyalists, it seems most everyone’s pretty happy. ‘[I]’m looking forward to returning to openai, and building on our strong partnership with msft,’ Altman tweeted. ‘[W]e are so back,’ Brockman exclaimed, sharing a selfie with his smiling team (who celebrated, according to the Information’s Erin Woo, by setting off a false fire alarm at OpenAI HQ). Twitch co-founder Emmett Shear is no longer interim CEO but is ‘deeply pleased by this result, after ~72 very intense hours of work,’ and is ‘glad to have been a part of the solution.’ Microsoft CEO Satya Nadella stated, ‘We look forward to building on our strong partnership and delivering the value of this next generation of AI to our customers and partners.’ (Brockman followed up with an expression of ‘incredible gratitude‘ toward Microsoft leadership.) Even MC Hammer, of all people, is psyched.
Still, as the Information’s Amir Efrati put it, ‘It’s not all roses.’ Neither Altman nor Brockman is retaining his old board seat. In fact, the makeup of OpenAI’s new board of directors is—Paul McCartney voice—very strange. Predictably, most of the mutineers who ousted Altman and Brockman are gone: Tasha McCauley and Helen Toner are no longer advising the company, and Ilya Sutskever, who ‘deeply regretted‘ his hefty role in this mess, is back at the company but not on the board. Weirdly enough, Quora CEO Adam D’Angelo is the only pre-Friday board member who’ll remain on what OpenAI characterized as its ‘initial board’ going forward, despite the employee letter insisting that that entire board step down. There, he’ll be joined by Google/Salesforce/Twitter alum Bret Taylor (who’s assuming Brockman’s former position as chairman) and by … Larry Summers, the august and/or chronically incorrect Harvard economist who predicted to Bloomberg last year that OpenAI’s ChatGPT bot ‘could be the most important general-purpose technology since the wheel or fire.’ In addition, Altman will voluntarily subject himself to an internal investigation over the reasons last week’s board members chose to boot him in the first place (the specifics of which we still don’t know). One can presume that D’Angelo’s continued board presence is a means by which to retain the input of an Altman opponent to begin with—although it also means that OpenAI’s leadership, for the time being, is dominated top to bottom by white men, a classic for Silicon Valley. (For what it’s worth, Nadella stated that Microsoft is ‘encouraged by the changes to the OpenAI board,’ claiming it’s a ‘first essential step on a path to more stable, well-informed, and effective governance.’)
For a tech press corps still mourning the end of high-octane boardroom-centric TV melodramas like Succession and Billions (as well as the close of several legal proceedings against former crypto CEOs Sam Bankman-Fried and Changpeng Zhao), it was a delicious five days of C-suite intrigue, with journalists and Twitter sleuths searching around for any clue they possibly could on just what the heck happened and why. Yet for all the relief that’s greeted Altman upon his return, this near-week of external investigation has surfaced some unflattering revelations.
As I wrote Monday, Altman’s ouster inflamed a long-simmering feud between two ideologies in the artificial intelligence space: on one side, the effective altruists deathly afraid of going too fast with A.I., and on the other, the ‘effective accelerationists,’ or e/accs, who believe we aren’t going fast enough. This mapped out somewhat imperfectly over the OpenAI sphere: the board members who were either explicit effective altruists (Tasha McCauley, Helen Toner) or thinkers at least sympathetic to their concerns (Ilya Sutskever, Adam D’Angelo), versus the inventors striving to usher in a superintelligent utopia (Altman, Brockman, their loyal staffers). Conflicting initial reports attributed Altman’s ouster to his disagreements with the board over A.I. safety, OpenAI’s much-celebrated role in developing high-powered A.I. tools, and OpenAI’s very purpose as a onetime nonprofit that was now pursuing full commercialism. As the board sparred with OpenAI investors/employees/executives over how to move forward, reporters uncovered key details around a long-festering internal battle sparked by the sudden, massive growth of ChatGPT. The New York Times reported Tuesday 1) that Sutskever ‘thought Mr. Altman was not always being honest when talking with the board’; 2) that OpenAI execs had been locked in a stalemate with board members over who should replace the directors who’d departed earlier this year, including Reid Hoffman, who was both an OpenAI founding member and a Microsoft board member; and 3) that Altman had been embroiled in conflict with board member Helen Toner in particular.
Let’s unpack that last one. Toner is a deep effective altruist, having worked exclusively in EA-specific institutions since March 2015. Per the Times’ report, Altman confronted Toner a few weeks ago over a paper she’d co-written for her current full-time employer, the EA-aligned Georgetown University Center for Security and Emerging Technology: ‘Mr. Altman complained that the research paper seemed to criticize OpenAI’s efforts to keep its A.I. technologies safe while praising the approach taken by Anthropic, a company that has become OpenAI’s biggest rival, according to an email that Mr. Altman wrote to colleagues.’ If you’re familiar with the lore, you’ll recognize Anthropic as the A.I. rival firm formed in 2020 by disgruntled former OpenAI leaders who disagreed with Altman over—wait for it—his approach to A.I. development. The OpenAI CEO viewed Toner’s paper as ‘dangerous’ for the company, especially since it was facing a government probe over the heaping amounts of data it uses to train its models; Toner defended her work as part of a legitimate discussion. Because of Altman’s claim that ‘any amount of criticism from a board member carries a lot of weight,’ his fellow executives, including Sutskever, pondered the possibility of booting Toner out altogether—but then Sutskever, who shared a somewhat ‘doomer’ outlook on the future of A.I., aligned himself with Toner to kick Altman out instead.
You’ll notice, then, that OpenAI’s current board has been purged of effective-altruist influence. (Per the CEO of A.I. company Perplexity: ‘Effective Altruism is officially dead.’) According to the Times, Toner was so devoted to the philosophy that she told OpenAI leaders, after firing Altman, that she’d rather have the company implode than have it continue on a more destructive path—in fact, she said, ‘that could be consistent with its mission’ of developing A.I. that ‘benefits all of humanity.’ (We don’t know much about the circumstances around the departure of EA-aligned board member Tasha McCauley, but according to Kara Swisher, the conflict with her went ‘deeper‘ thanks to her ‘apocalyptic’ viewpoints.) If you take Toner’s word for it, her actions were merely in line with the OpenAI charter, whose tenets are evidently drilled into everyone affiliated with the firm. At the same time, what she may not have counted on was the support Altman inspired from the company’s employees. As Karen Hao reported for the MIT Technology Review back in 2020, ‘OpenAI, for all its noble aspirations, is obsessed with maintaining secrecy, protecting its image, and retaining the loyalty of its employees.’
This speaks to another important factor in the current battle: Altman’s own ruthlessness, a core aspect of his leadership belied by his soft-spoken public appearances and requests for government officials worldwide to please regulate A.I. The timeline of his control over OpenAI, from its 2015 founding through its November 2023 rebirth, is chock-full of back-and-forth conflicts that threatened Altman’s visions and goals for OpenAI, all ending with his would-be challengers packing their bags and fleeing the premises. A compact overview:
• In 2018, when OpenAI co-founder and co-chair Elon Musk complained that the firm was falling behind in the A.I. competition, he proposed fully taking over the venture himself. Altman and Brockman rejected this, causing Musk to leave and pull all his planned funding for the then-nonprofit—and allowing Altman to assume full OpenAI control.
• To make up for the Musk-caused cash gulf, Altman decided in 2019 to establish a for-profit arm of OpenAI that could raise more outside funding at a faster pace, allowing it to remain competitive with Big Tech’s A.I. adventures. However, some of OpenAI’s more EA-aligned members were alarmed by this capitalistic turn; the whole reason OpenAI started as a nonprofit was so its A.I. ventures wouldn’t be dangerously motivated by profit. Per the aforementioned Times report, these dissenters themselves asked the board to force Altman out before resigning to form the EA-supported rival A.I. firm Anthropic.
• Ironically, Anthropic now enjoys ample funding from Big Tech giants like Amazon, as well as an app partnership with Adam D’Angelo’s Quora. Oh, and also: In the past few days, some OpenAI board members approached Anthropic about possibly merging the two companies, a proposal firmly rejected by Anthropic leadership.
If you’re keeping track at home, then, the most recent events mark the third time since 2018 that someone at OpenAI has attempted to usurp Sam Altman’s power, only to give up, get out, and, in the earlier cases, establish their own competitors. (Musk, you may recall, is currently trying to make his own painfully unfunny chatbot a thing.) It’s all just more confirmation, perhaps, of Ellen Huet’s description of the CEO in a recent Bloomberg profile: having ‘a reputation as ambitious, cunning, even Machiavellian.’ No wonder, then, that so many OpenAI leaders and workers have stayed by his side since Friday, and that so few others in his orbit ever attempted to follow the path of Musk, the Anthropic founders, and Toner—betting against Altman’s position, and betting that they can take his place. This week may demonstrate yet again the folly of this approach. Why else would Ilya Sutskever help to oust Altman only to get cold feet and publicly regret his decision?
The thing is, Altman is not entirely unsympathetic to the effective-altruist outlook on A.I. He’s spoken time and again about the need to be responsible with powerful A.I., and he’s been influenced by many of the rationalist/longtermist philosophers whose A.I. musings have helped to influence the effective-altruist approach. But here’s another thing: Altman has clearly determined that he is the right proper steward of the A.I. future, and no one is going to cross him in this path—especially not now, when he’s become such a public face for A.I. Nothing open about that.
Reference: https://slate.com/technology/2023/11/sam-altman-rehired-openai-effective-altruism.html
Ref: slate
MediaDownloader.net -> Free Online Video Downloader, Download Any Video From YouTube, VK, Vimeo, Twitter, Twitch, Tumblr, Tiktok, Telegram, TED, Streamable, Soundcloud, Snapchat, Share, Rumble, Reddit, PuhuTV, Pinterest, Periscope, Ok.ru, MxTakatak, Mixcloud, Mashable, LinkedIn, Likee, Kwai, Izlesene, Instagram, Imgur, IMDB, Ifunny, Gaana, Flickr, Febspot, Facebook, ESPN, Douyin, Dailymotion, Buzzfeed, BluTV, Blogger, Bitchute, Bilibili, Bandcamp, Akıllı, 9GAG