More high-profile departures at OpenAI – this time John Schulman, co-founder and chief research officer, Greg Brockman, president and co-founder, and Peter Deng, head of product.
This latest restructuring got here just months after a serious shakeup in OpenAI's AI security team, further exacerbating the corporate's recent pattern of instability.
Perhaps most shocking is Schulman's intention to hitch OpenAI competitor Anthropic, best known for its Claude series of language models.
Anthropic and OpenAI models are battling it out at the highest of the machine learning leaderboards, with recent releases competing directly with one another (e.g. GPT-4o with Claude 3.5 Sonnet, GPT-4o mini and Claude Haiku, etc.).
This shouldn’t be an isolated case. OpenAI has brought quite a lot of drama to the AI sector in recent months, with quite a few high-profile departures and the leadership debacle in late 2023 that saw CEO Sam Altman fired and rehired inside days.
Is all this an indication of a deeper, systemic problem in the corporate, a coincidence or a storm in a teacup?
Whatever the reply, it doesn’t exactly paint an image of stability at an organization that is taken into account a pioneer in the long run of AI research.
John Schulman: Co-founder and Chief Research Officer
Perhaps essentially the most surprising departure is that of John Schulman, certainly one of the co-founders of OpenAI and a highly respected figure in AI research.
Schulman announced his decision to go away OpenAI and join rival AI company Anthropic, citing a desire to focus more on AI direction—the critical challenge of ensuring AI systems behave in ways which are secure and helpful to humanity.
Previously, OpenAI executives Jan Leike and Ilya Sutskever and various other employees had made similar statements that raised doubts inside the company concerning the ability to handle the safety of AI.
Schulman explained to X:
“Today I announced to my OpenAI colleagues that I actually have made the difficult decision to go away OpenAI. This decision stems from my desire to deepen my give attention to AI direction and start a brand new chapter of my profession where I can return to hands-on engineering work. I actually have decided to pursue this goal at Anthropic, where I imagine I can gain latest perspectives and conduct research alongside people who find themselves deeply engaged within the topics that interest me most.”
I shared the next note with my OpenAI Colleagues today:
I made the difficult decision to go away OpenAI. This decision stems from my desire to deepen my give attention to AI direction and begin a brand new chapter of my profession where I can return to hands-on technical work. I actually have decided…
— John Schulman (@johnschulman2) 6 August 2024
Schulman's departure is especially significant given his instrumental role in developing a few of OpenAI's most influential technologies, including advances in reinforcement learning.
Altman gave Schulman a positive farewell, posting: “Thank you for all the things you've done for OpenAI! You are an excellent researcher, a deep thinker about products and society, and most of all, a terrific friend to all of us. We will miss you terribly and make you happy with this place.”
Thank you for all the things you’ve gotten done for OpenAI! You are an excellent researcher, a deep thinker about products and society and most of all you’re a terrific friend to us all. We will miss you terribly and you’ll make this place proud.
(I first met John in a restaurant in Berkeley in …
-Sam Altman (@The same) 6 August 2024
Greg Brockman: President and Co-Founder (prolonged vacation)
Greg Brockman, president of OpenAI and one other co-founder, is taking an prolonged leave of absence until the top of the 12 months.
Brockman was a very important public face for the corporate and played a key role in its growth and strategic direction.
“I'm taking a break until the top of the 12 months. Time to chill out for the primary time since co-founding OpenAI 9 years ago. The mission is way from complete; we still need to construct secure AGI,” Brockman posted on X.
I'm taking a break until the top of the 12 months. Time to chill out for the primary time because the company was founded. OpenAI 9 years ago. The mission is way from complete; we still need to construct a secure AGI.
— Greg Brockman (@gdb) 6 August 2024
Although Brockman has described his vacation as a time without work, the timing is clearly unusual. Is he fascinated by his place at OpenAI?
One could definitely make that assumption considering he’s certainly one of not only two, but three key people leaving the corporate.
If he were a detailed ally of OpenAI and Altman, he probably would have delayed announcing the news, even though it may not have made a difference.
Peter Deng: Product Manager
The third departure is Peter Deng, who joined OpenAI in 2023 after holding senior roles at firms similar to Meta, Uber and Airtable.
That's a brief tenure, and Deng's post was crucial at a time when the corporate was rapidly commercializing its AI technologies and facing strong competition.
A pattern of outputs
This shift doesn’t occur in a vacuum. In May 2024, OpenAI's “Superalignment” team, which was alleged to align AI safety with moral and societal values, appeared to implode:
- Ilya Sutskever – Co-founder and Chief Scientist of OpenAI
- Jan Leike – Head of Alignment and Super-Alignment Management
Leike's departure was particularly controversial. He stated publicly: “In recent years, safety culture and processes have lagged behind shiny products.”
Sutskever later founded his own company, Safe Superintelligence Inc., which could absorb any talent leaving OpenAI.
The May exodus was not limited to those two individuals. In recent months, no less than five other key members of the Super Alignment team, including Daniel Kokotajlo, Leopold Aschenbrenner and Pavel Izmailov, have either quit or were reportedly forced to go away.
And let's not forget last 12 months's leadership debacle, by which Altman survived a boardroom coup and was kicked out of the corporate, only to rejoin days later after an uproar amongst his team.
The employees supported Altman and the corporate on the time and published the statement “OpenAI is nothing without its people” – a sentence that’s now taking its revenge.
Growing challenges for OpenAI
And so as to add fuel to the fireplace, these high-profile departures come amid a series of challenges for OpenAI that don’t have anything to do with leadership.
First, GPT-5 is nowhere to be seen. GPT-4o will certainly fall back into the pack within the meantime, with OpenAI to TechCrunch“We don’t plan to announce our next model at DevDay” and “We will focus more on informing developers about available products and showcasing stories from the developer community.”
OpenAI published end of May that it’s busy training its next frontier model (presumably GPT-5), the success of which is now absolutely critical to the corporate's status within the AI sector.
At least the much-hyped voice communication features of GPT-4o are introduced, but this was not the announced groundbreaking release In OpenAIDemo of.
In addition, it was not way back The information revealed that OpenAI's costs are soaring. This 12 months, the corporate reportedly spent about $7 billion on model training and $1.5 billion on personnel. This 12 months, it could incur losses of $5 billion.
Generative AI is hugely expensive, and there are doubts as as to if OpenAI can truly maintain the momentum needed to push the industry's pace forward because it once did.
In addition, Elon Musk has filed a brand new lawsuit against OpenAI and Altman in a California federal court under the Racketeer Influenced and Corrupt Organizations Act. It is a fair stronger lawsuit than the unique, with 15 counts as an alternative of 5.
Musk argues that OpenAI's pursuit of profits is a breach of the contract he and co-founders Altman and Brockman signed once they founded the corporate.
There has all the time been trouble within the tech paradise of Silicon Valley, but OpenAI is currently a curious case.
Brockman's moves at the top of the 12 months might be very telling as to how these recent departures impact the long run of the corporate.
Whether it will impact OpenAI’s Next Frontier model or other products stays to be seen.
Altman will definitely do all the things he can to proceed working toward the subsequent milestone.
