It's been a strange week at OpenAI…
First, there has been a series of high-profile departures that have longtime AI watchers doubting all is well at the company.
Second, CEO Sam Altman has been publishing and engaging in cryptic posts online that refer to the company's flagship “Strawberry” project that was leaked earlier this year.
Together, these developments could signal something big is in the works at the world's most famous AI lab.
What is really going on here?
I got the scoop from Paul Rutzer, founder and CEO of the Marketing AI Institute. Episode 110 of Artificial Intelligence Show.
The Great OpenAI Exodus?
First, let's look at departures. Who's left here:
- Greg Brockman, president and co-founder of OpenAI, announced that he is taking an “extended absence” until the end of the year.
- John Shulman, another co-founder and key AI leader, left to join rival company Anthropic.
- Peter Deng, a product leader who joined OpenAI from Meta last year, has also left.
The exodus has led some industry observers to question OpenAI's progress. As AI researcher Benjamin D. Kracker put it at X: “If OpenAI is on the brink of AGI, why do prominent people keep leaving?”
“All we can really do at this point is speculate,” Rutzer says. “It's complicated, and I think we just have to let it play out. I have a lot of unanswered questions about Greg's time off.”
It's entirely possible that Brockman's layoff is an extended one.
But it is difficult. No To speculate, given what Sam Altman was doing online with these announcements, that something big was going to happen.
The mystery surrounding Project “Strawberry”.
On August 7, Sam Altman posted a photo of strawberries from his garden on X, a possible nod to months of rumors about an internal project at the company called Strawberry. Many speculate that Strawberry refers to a major advance in advanced reasoning that will be present in the next model of OpenAI.
A few hours later, an anonymous account on X posts the following with strawberry emojis as its profile name, with which Sam Altman himself immediately engages:
“Level 2,” Rutzer says, referring to the five-level scale that OpenAI created to track its progress toward artificial general intelligence (AGI). which was first reported in July 2024..
Level 2 on this scale is the next step beyond the AI models we have: models that can reason and engage in human-level problem solving.
The mysterious account, which includes a profile picture of Joaquin Phoenix from the AI movie “Her,” has since flooded Twitter with AI-related posts, with many citing Altman's strawberry tweet and this Says something big is about to be released.
Some AI watchers speculated that the account was being run by an OpenAI insider to tease a bigger launch. Others believe it could be an advanced form of self-accounting AI.
“I guess it's some form of AI,” Rutzer says. “I think there's a human here, but I don't think there's a human managing it.”
Whose model is this? Rutzer doesn't know. But he suspects that this strange series of events means that OpenAI's next step is about to unfold.
“I think the construction of the next model is done, and they are just finalizing the timing and plans for the release of this model.”