Understanding the OpenAI-Chaos
Plus: Unviral Bin Laden.
My latest posts for the german expert community platform Piqd, machine translated with ChatGPT and edited for some punch. This November, besides some stuff i already posted here, i wrote about the upheaval at OpenAI and it’s consequences, and the Bin Laden-letter which went not very viral on TikTok.
GOOD INTERNET is a reader supported online mag. If you like what i do here, you can support this thing by upgrading your subscription to a paid plan or use one of the other support options you can find at the bottom of this issue.
Understanding the Chaos at OpenAI
Over the weekend, we witnessed a severe earthquake in the AI industry, and the consequences are still out.
The events unfolded as follows:
On Friday afternoon, the board of OpenAI orchestrated fired CEO Sam Altman in a company-wide coup led by Chief AI Researcher Ilya Sutskever. Mira Murati takes over Altman's position and co-founder Greg Brockman resigns in solidarity. Microsoft, the largest investor, was surprised by these events and informed only minutes before the announcement.
On Saturday, rumors circulated about the CEO's potential return and the board's resignation. Investors, employees, and managers within the company had rallied in support of Altman and Brockman. Altman demands a restructuring of the board and issues an ultimatum.
Negotiations break down, Twitch CEO Emmett Shear becomes the new interim CEO.
Microsoft hires Sam Altman and Greg Brockman for a new AI research team.
To make sense of the chaos involving some of the leading minds in the AI industry, one must first understand the corporate structure of OpenAI.
OpenAI Inc. was founded in 2015 as a non-profit with the goal of conducting AI research focused on safety and AI alignment "for the benefit of all humanity," free from profit motives and commercial interests. Concurrently, there is Open AI LLC, a company developing commercial applications. Microsoft, as the largest funder, is a minority owner of this commercial arm.
Sam Altman pretty much represents OpenAI's commercial arm, advocating for faster product development and commercial exploitation of the tech, especially after the global success of ChatGPT. In contrast, Ilya Sutskever expressed safety concerns and envisioned the development of General AI (AGI) which he expects to arrive soon. Within the company, two "tribes" emerged — one focused on commercial exploitation, the other adhering to the company’s original goal of safe AGI development aligned to human values. The coup resulted from an internal power struggle between these two factions, and while it may superficially appear that the Safe-AGI faction won, the overall picture seems to be the opposite.
In a memo seen by the New York Times, the board stated:
“The board firmly stands by its decision as the only path to advance and defend the mission of OpenAI,” said the memo, referring to Mr. Altman’s removal from the company on Friday. It was signed by each of the four directors on the company’s board; Adam D’Angelo, Helen Toner, Ilya Sutskever, and Tasha McCauley.
“Put simply, Sam’s behavior and lack of transparency in his interactions with the board undermined the board’s ability to effectively supervise the company in the manner it was mandated to do,” the memo said.
On one hand, the board's removal of Altman seems to strengthen its role in the safety-focused original company mission.
On the other hand, the tumultuous events are likely to deter many investors and funders, jeopardizing the primary goal of AGI development. It is unknown whether OpenAI is already profitable or if the ongoing costs for servers and GPUs are eating up the substantial monthly revenue of $80 million, especially given the considerable user numbers. OpenAI, as a nonprofit, relies on investor funding, and its future is uncertain after the debacle.
Furthermore, Microsoft's hiring of Altman and Brockman indicates their continued aggressive expansion of AI technology, whether with products from OpenAI or its competitors. Microsoft right now integrates AI into all its products, from the Windows operating system to Office to Github Copilot. The market power behind Microsoft's AI endeavors is immense, especially with their Azure cloud computing platform, running OpenAI's products, experiencing a 29% revenue increase in the third quarter of 2023 alone.
Just before his ousting on Thursday, Altman announced new breakthroughs in AI development at the APEC CEO Summit, increasing pressure on the parts of the company prioritizing AI safety. With Altman now leading a new AI research team at Microsoft, who are also shareholders in OpenAI LLC, they are in a prime position, controlling not only OpenAI's cash flow as the largest investor but also having Altman and Brockman as charismatic leaders on board. This development is likely to further intensify the split within OpenAI.
On its surface, the apparent final removal of Sam Altman looks like a victory for the safe development of AI. In reality, Microsoft is the true winner of this coup, fueled by massive revenue from OpenAI's commercial arm, and they will significantly expand their AI development, whether with OpenAI products or a possible new open-source strategy, similar to Facebook/Meta. As a critic of open-source AI development, these developments make me raise an eyebrow or two.
For a good measure, here are movie recommendations for this drama: Knives Out and, of course, Game of Thrones. Automatic AI popcorn included.
Drama Updates
Hundreds of OpenAI employees threaten to resign and join Microsoft: “In a letter to OpenAI’s board that was reported on this morning by Wired and journalist Kara Swisher, more than 500 current OpenAI staffers say that ‘Microsoft has assured us that there are positions for all OpenAI employees at this new subsidiary should we choose to join.’ The letter says the OpenAI employees will leave if the board does not reinstate Altman and Brockman and then resign. Swisher reports that there are currently 700 employees at OpenAI and that more signatures are still being added to the letter.“
OpenAI's board approached Anthropic CEO about top job and merger
OpenAI in ‘Intense Discussions’ to Unify Company, Memo Says: “OpenAI Vice President of Global Affairs Anna Makanju said in a memo to staff Monday that the company is in “intense discussions” to unify its divided staff, according to a statement reviewed by Bloomberg. ‘We can assure you that our number one goal remains to reunify OpenAI and discussions are actively ongoing,’ Makanju said in the memo. She also said that the company is in touch with Altman, new CEO Emmett Shear and the company’s board, ‘but they are not prepared to give us a final response this evening.’”
Sam Altman is still trying to return as OpenAI CEO:: “Sam Altman’s surprise move to Microsoft after his shock firing at OpenAI isn’t a done deal. He and co-founder Greg Brockman are still willing to return to OpenAI if the remaining board members who fired him step aside, multiple sources tell The Verge (…) Microsoft CEO Satya Nadella appeared on CNBC and Bloomberg TV. When asked directly by CNBC’s Jon Fortt if Sam Altman and OpenAI’s staffers would join Microsoft, Nadella said ‘that is for the OpenAI board and management and employees to choose.’“
Altman Sought Billions For Chip Venture Before OpenAI Ouster: “Sam Altman was actively working to raise billions from some of the world’s largest investors for a new chip venture (…) The OpenAI chief executive officer planned to spin up an AI-focused chip company that could produce semiconductors that compete against those from Nvidia Corp., which currently dominates the market for artificial intelligence tasks. Altman’s chip venture is not yet formed and the talks with investors are in the early stages, said the people, who asked not to be named as the discussions were private.“
OpenAI board stands firm in face of staff revolt over Sam Altman’s ouster: “By the end of Monday, 747 out of 770 OpenAI employees had signed a letter threatening to quit and join Microsoft (…) Meanwhile, venture capitalists backing the generative artificial intelligence start-up swung behind staff demands and were exploring legal measures to force the board to reverse course“.
And the finale: Sam Altman to return as CEO of OpenAI: “Sam Altman will return as CEO of OpenAI, overcoming an attempted boardroom coup that sent the company into chaos over the past several days. Former president Greg Brockman, who quit in protest of Altman’s firing, will return as well. The company said in a statement late Tuesday that it has an ‘agreement in principle’ for Altman to return alongside a new board composed of Bret Taylor, Larry Summers, and Adam D’Angelo. D’Angelo is a holdover from the previous board that initially fired Altman on Friday. He remains on this new board to give the previous board some representation, we’re told.“
OpenAI researchers warned board of AI breakthrough ahead of CEO ouster, sources say: “Some at OpenAI believe Q* (pronounced Q-Star) could be a breakthrough in the startup's search for what's known as artificial general intelligence (AGI), one of the people told Reuters.“
Unviral Bin Laden
People are consuming less and less news on social media, and the numbers of news consumption on major social media platforms are all showing a downward trend, with one notable exception: TikTok.
In just three years, the number of people watching news on the video platform has quadrupled. A year ago, I wrote about how TikTok had surpassed the former top viral source, Twitter, as the "Number 1 meme source." Viral news now happens on TikTok, making it crucial, especially for journalists, to accurately assess the virality of individual pieces of information on the platform.
Since yesterday, TikTok has been deleting videos under the hashtag #lettertoamerica, where TikTok users recontexualize a twenty-year-old letter from Osama Bin Laden concerning recent developments in Israel and Gaza. The New York Times reports a total of 14 million views for the hashtag, spread across 274 videos created before the viral backlash against the alleged internet phenomenon. This equates to approximately 51,094 views per video. While these numbers may seem high, they are minuscule compared to genuinely viral TikTok videos.
In a piece about the new vaporized internet and the White House's overreaction to an alleged imbalance between the use of pro-Palestine and pro-Israel hashtags on TikTok, Ryan Broderick describes the actual virality numbers on the platform. The most successful video in October had 37 million likes and 313 views. Keep in mind, this is for a single video. The hashtag #Halloween garnered 20 billion views within three months. Compared to such genuinely viral phenomena, 14 million views spread across 274 videos are peanuts.
In piece on Slate, Scott Nover rightly points out that it's not the videos themselves that went viral — but the backlash against them:
One prominent Twitter figure’s outraged post about the videos, which included a supercut of them, racked up 32 million views on X. Flipping through TV channels on Thursday, I noticed the story on several different news broadcasts—every anchor and reporter was disgusted and wanted to say so. The Biden administration even responded Thursday. “No one should ever insult the 2,977 American families still mourning loved ones by associating themselves with the vile words of Osama bin Laden,” White House spokesperson Andrew Bates wrote on X while sharing the CNN article.
The bigger picture here is this: one of the causes of the outrage cascades in the past decade has been major media outlets exploiting individual opinions on social media, especially conservative outlets like the Telegraph, reporting on stuff like three barely noticed tweets from leftist LGBTQ activists or some minority opinion by some campus unions, fuming the flames of a global backlash to social justice politics. There sure are some problems with identitarian leftwing politics, but media outlets exploiting them based on a few extreme tweets did not help.
Now, this practice morphed into full fledged political agenda setting. Ryan Broderick on this:
this is yet another example of how everything we see about TikTok is just being filtered through a media and political apparatus that doesn’t know how it works and has no issue cherry-picking random nonsense out of the app to fit whatever agenda they subscribe to. As New York Mag’s John Herrman wrote, “[The Bin Laden letter] became popular much in the way that posts from accounts like @LibsofTikTok do, by giving users license, via a video ripped from elsewhere, to drop their inhibitions and rage out a little bit.” Which is the real takeaway here.
Baseless generational in-fighting, aging millennials who refuse to accept the new status quo of the internet, easily monetizable rage bait, lazy TikTok trend reporting, and bad faith political actors swirled together to create a perfect storm this week. We have invented a version of TikTok that simply does not exist and now many people in power are ready to tear apart the foundation of internet to prove it does.
The responsibility of journalism in the 21st century includes contextualizing and accompanying the dynamics and qualities of virality — or to even refrain from reporting on what a few isolated accounts on TwiX, or what a meager 274 videos on TikTok think. Not every stupid opinion on the internet needs to be amplified and made viral by the press, even though it is a guaranteed click machine because of its extreme and/or extremely foolish nature.
There is nothing more emotionally charged on the internet than being able to point the finger at the alleged multitude of stupid idiots, of which there are plenty, i can tell you that — with the absolute certainty that you ofcourse belong to the righteous, good and decent ones. The fools are always the other, and it is all too easy for editorial teams to pluck the absolutely unspeakable from the sheer sea of opinions and sell it as a phenomenon — such as the uninteresting TikTok chatter from a few kids who think that what a dead islamofascist terrorist says somehow fits well with the Gaza conflict.
Scott Nover succinctly concludes:
A small group of nobodies on TikTok saying dumb shit is not a viral trend that necessitates mass hysteria in response. This incident is only news because, well, it became news.