OpenAI’s drama was swept under the rug, but apparently, someone has some new dirt. A former OpenAI board member spoke out about why Sam Altman was fired from his post temporarily.
To refresh your memory, there was a massive coup that happened back in November 2023 that saw the company’s CEO Sam Altman fired. It was a dramatic weekend that saw Microsoft nearly hire Altman, a potential mass walkout of the employees, several interim CEOs, and a new non-profit board within the company. We also got a hot tip on something called Q*, something that researchers feared would endanger the human race.
All of this drama eventually cooled just a few days after it began. As the next work week started, most of the storm had passed, and we all went on with our lives. But, there were still questions in the air. What was Sam Altman not being completely candid about? Is the current tension within the company a product of the coup? More importantly, what is Q*, and has the company already launched it?
Former board member speaks about why Sam Altman was fired
Unfortunately, most of the questions in the air won’t touch the ground anytime soon, but the comments from former board member Helen Toner show us that the drama might be far from over. She recently spoke on The TED AI Show podcast with Bilawal Sidhu. The 41-minute episode was mostly about AI in general, but at the beginning of the podcast, Toner spilled some tea about why the board agreed to fire Altman.
The non-profit board that Toner was on was charged with the duty of making sure that the company stayed on track with its mission to benefit the public over profits. Much of the drama with the coup surrounded Altman’s interactions with the board, according to her account. She said that he wasn’t entirely honest about certain important details, and she said that he outright lied about certain things as well.
She stated that his actions made it hard for the board to do its job. It relied on having solid information from him about the company and its inner workings to be able to properly carry out its mission, but he seemed to withhold a lot of information from them.
ChatGPT’s launch and the OpenAI startup fund
A good example she gave was the fact that the board wasn’t informed about the launch of ChatGPT beforehand. “We learned about ChatGPT on Twitter,” she said. That must have felt like a slap to the face, but things go deeper than that.
Another example she gave was the fact that Sam failed to mention that he owned the OpenAI Startup fund. As CEO, you’d expect him to have some sort of interest in the company, but Sam said that he was an independent board member and that he didn’t have any financial interest in the company. Well, it appears that this wasn’t true, allegedly.
Safety protocols
Next up, according to Toner, Altman wasn’t perfectly honest about the number of safety protocols in place within the company. “On multiple occasions, he gave us inaccurate information about the small number of formal safety processes that the company had in place.”
This is something that we can’t 100% confirm, but recent reports about the company do lend some weight to it. According to a former researcher Jan Leike, “Over the past years, safety culture and processes have taken a backseat to shiny products.” Apparently, OpenAI had been reducing the number of resources going to its previous safety team over time. Not only that, but the company recently established a new safety team. So, it’s obvious that OpenAI’s approach to its products’ safety isn’t stable. It’s believable that the company wouldn’t have too many safe protocols in place.
The paper
Lastly, Toner referred to a research paper that she co-authored targeted at policymakers in the Fall of 2023. The paper itself seemed to criticize OpenAI’s practices and favor Anthorpic’s. Obviously, this is something that didn’t sit well with Altman. It led to a perfect storm of drama for the company, as this was when it was being investigated by the FCC about how it gathers and uses data. So, the company was not looking good in the public eye at that point.
After the paper was published, according to Toner, Altman started going behind her back and telling stories to the board in order to have her kicked off. However, it seemed that the board wasn’t quite on his side. Toner said that, in late October last year, the board was “talking pretty seriously” about whether to fire him or not.
Trust issues
So, going by Toner’s account, it seems that Altman just had a thing about lying to get his way. Again, we can’t completely confirm this, but reports that came out around the time of the drama corroborate these claims. After Sam Altman was ousted, the board came together and said that he wasn’t “consistently candid” with the board.
So, it seemed that the board got wind of how he was behaving, and that was one of the reasons for his firing. “The end effect was that, after years of this kind of thing, all four of us who fired him came to the conclusion that we just couldn’t believe things that Sam was telling us.” She followed up by saying that it’s a “completely unworkable place to be in as a board,”.
Casting a harsher light on Sam, other executives have come forward to the board to express that they couldn’t trust him either. Also, they mentioned that he created a toxic atmosphere and thought that he wasn’t the right person to lead the company.
The mass walkout
If you remember, during the drama that happened at OpenAI, numerous employees threatened to walk out of the company if Sam wasn’t reinstated as the CEO. Toner mentioned that there was a lot of misreporting on that. She said that there were a few things at play during that scenario. Firstly, the workers seemed to be presented with only two options: either they rally to bring Sam back or they risk OpenAI being completely destroyed.
Next, and this does NOT paint Altman as a good person, Toner stated that many people were afraid of Sam retaliating against them. If Altman did create a toxic work environment, it’s natural to believe that many people didn’t want to get on his bad side.
Past jobs
One of the biggest bombshells of this whole situation has to do with the previous jobs that Altman allegedly held… and was fired from. Again, we can’t confirm this information completely, so we’ll need to wait for more information to do so. However, we’re going by Toner’s account,
According to her, Altman has a bad track record of being fired from other posts. Apparently, he was fired from his job at Y Combinator. This is a company that helps startups get off of the ground. It has a long list of notable companies under its belt including Twitch, DoorDash, airbnb, and more.
We don’t know the reasons why he was fired, but there’s another job that seemingly ousted Altman. Previously, he had a startup called Loopt. According to Toner, the management team there complained to the board asking for him to be fired. They complained about his “deceptive and chaotic behavior.” If this is true, then it falls in line with what got him fired from OpenAI.
OpenAI’s response
OpenAI’s board chair Bret Taylor gave a response to Toner’s words, stating that the company is “disappointed that Ms. Toner continues to revisit these issues.” It continued to state that “the prior board’s decision was not based on concerns regarding product safety or security, the pace of development, OpenAI’s finances, or its statements to investors, customers, or business partners.” Unfortunately, outside of denying the claims, the statement didn’t give any additional insight as to why he was fired. So, it’s a bit tough to take Taylor at his word. “Additionally, over 95 percent of employees, including senior leadership, asked for Sam’s reinstatement as CEO and the resignation of the prior board.”
The statement finishes by stating that, “Our focus remains on moving forward and pursuing OpenAI’s mission to ensure AGI benefits all of humanity.” Let’s not mince words here, the statement provided by Taylor seems to be a very scripted and PR-focused statement. So, it’s a little hard to take it at face value.
Who to believe?
Helen Toner provided a ton of inside baseball that the public didn’t know before, and Bret Taylor’s statement didn’t outright deny any of the information about the toxic work environment, lies, previous jobs, or deceptive practices. So, it’s much easier to take Toner’s word over anything coming from OpenAI.
Also, we have to remember that Helen Toner is not currently associated with OpenAI while Bret Taylor is. What is Taylor going to do, talk smack about the CEO of the company he’s currently on the board for? There’s much more incentive for him to sweep details under the rug and defend OpenAI. However, since Toner’s no longer associated with OpenAI, we can expect her comments to be more truthful.
In all honesty, there’s a possibility that Toner herself isn’t being 100% honest, as we’re waiting for more information on this drama to come out. However, the genie is out of the bottle at this point. We should hopefully have more people come forward and spill more tea.
2024-05-30 15:09:20