On August 6, John Schulman, one of the founders of OpenAI, announced at X that he would be leaving OpenAI and joining AI startup Anthropic, one of OpenAI's competitors.

In the post, Schulman said that he chose to leave OpenAI in order to delve into AI alignment and return to actual technical work. By AI alignment, we mean guiding the behavior of AI systems to align with human values and interests.

Schulman stated, "I decided to pursue this goal at Anthropic where I believe I can gain new perspectives and do research with people who are deeply involved in the topics that interest me most."

Schulman emphasized, "It's important to be clear that my departure is not due to a lack of support for AI alignment research at OpenAI. On the contrary, company leaders have been committed to investing in this area. My decision is a personal one, based on how I want to focus my efforts in the next phase of my career."

Schulman also expressed gratitude to the other founders (e.g., CEO Sam Altman) and colleagues, stating, "Together, we have built an extraordinary and unprecedented company with a mission of public good. I'm confident that OpenAI and the team I'm part of will continue to thrive even without me."

Under the post, ChatGPT's official account replied, "Thank you for all you've done to build the machine that models ChatGPT."

Sam Altman, on the other hand, commented, "Thank you for all you do for OpenAI! You've been a brilliant researcher, a deep thinker about the product and the community, and most importantly, a great friend to all of us. We will miss you terribly and make you proud of this place."

Several members of OpenAI's security team have announced their departure so far this year, most notably Jan Leike, an AI security researcher at OpenAI who also joined Anthropic after coming out of OpenAI.

After Jan Leike's departure, Schulman became the head of OpenAI's alignment science team, also known as the "post-training" team. Schulman is also a member of OpenAI's recently formed Safety Committee. It's unclear who will replace Schulman.

Schulman is expected to work alongside Jan Leik, both of whom have a strong background in AI safety.

With Schulman's departure, OpenAI's 11 original founders are down to just three: Sam Altman, OpenAI's CEO; Greg Brockman, its president; and Wojciech Zaremba, its head of language and code generation.

And, on the same day that Schulman announced his departure from OpenAI, Brockman also posted on X that he would be taking a sabbatical until the end of the year. This is Brockman's first sabbatical since he founded OpenAI as a co-founder nine years ago. In the post, Brockman also said, "Our mission is far from complete and we still need to build secure AGI."

Brockman had been a key ally of Altman last year when Sam Altman, the CEO of OpenAI, was ousted. Brockman also resigned as chairman of the board of directors as a show of support for Brockman.

The fact that both co-founders are stepping down from the front lines of the company may indicate that the "fight" within OpenAI is not over. The debate over AI safety has gradually subsided with the departure of a number of "patriarchs".

Anthropic, which was founded in 2021 by an elite seven-person team led by former OpenAI veteran Dario Amodei, seems to be the winner of the OpenAI staffing change.

Anthropic has stated that the company's goal is to build a reliable, explainable, and controllable AI system that has human interests at its center. It can be said that Anthropic emphasizes more on the concern for AI safety, which is also a major reason why the company has attracted several OpenAI employees to jump ship.

Currently, Anthropic is considered one of OpenAI's strong rivals. As the competition between the two sides becomes more and more intense, this battle for employees may be far from over, and let's wait and see who will eventually be better.