Anika Begay
A ChatGPT chat screen on a smartphone seen in the Brooklyn neighborhood of New York, USA, Thursday, March 9, 2023. ChatGPT has made it easier to write computer code and cheat on homework. Soon, it could make email scams a breeze, according to British cybersecurity firm Darktrace Plc.
Gabby Jones | Bloomberg | Getty Images
OpenAI co-founder John Schulman said in a Monday blog post that he would be leaving the Microsoft-backed company and join Anthropic, an AI startup with funding from Amazon.
The move comes less than three months after OpenAI disbanded a superalignment team that had been focused on trying to ensure that people can control AI systems that surpass human capabilities in many tasks.
Schulman had been a co-leader of OpenAI’s post-formation team that refined AI models for the ChatGPT chatbot and a third-party developer programming interface, according to a bio on his website. In June, OpenAI said Schulman, as its head of alignment science, would join a safety and security committee that would advise the board. Schulman has been working at OpenAI since receiving a Ph.D. in computer science in 2016 from the University of California, Berkeley.
“This move is driven by my desire to deepen my focus on AI alignment and begin a new chapter in my career where I can return to hands-on technical work,” Schulman wrote in the social media post.
He said he wasn’t leaving because there was a lack of support for new work on the topic at OpenAI.
“On the contrary, company executives have been very committed to investing in this area,” he said.
Superalignment team leaders Jan Leike and company co-founder Ilya Sutskever both left this year. Leike joined Anthropic, while Sutskever said he was helping found a new company, Safe Superintelligence Inc.
Since OpenAI staffers founded Anthropic in 2021, the two young San Francisco-based companies have been competing to have the highest performing generative AI models that can process human-like text. Amazon, Google, and Meta have also developed large language models.
“So excited to be working together again!” Leike wrote in response to Schulman’s message.
Sam Altman, co-founder and CEO of OpenAI, said in a blog post that Schulman’s perspective inspired the startup’s initial strategy.
Schulman and others chose to leave after the board ousted Altman as chief last November. Employees protested the decision, prompting Sutskever and two other board members, Tasha McCauley and Helen Toner, to resign. Altman was reinstated, and OpenAI hired additional board members.
Toner said in a podcast that Altman had provided the board with incorrect information about the company’s “small number of formal security procedures.”
Law firm WilmerHale found in an independent review that the board was not concerned about the product’s safety when it excluded Altman.
Last week, Altman said on X that OpenAI “has partnered with the US AI Safety Institute on an agreement where we will provide early access to our next baseline model so we can work together to advance the science of AI assessments.” Altman said OpenAI is still committed to setting aside 20 percent of its computing resources for safety initiatives.
Also on Monday, Greg Brockman, another OpenAI co-founder and its chairman, announced that he would be taking a sabbatical for the remainder of the year.
CLOCK: OpenAI Announces Search Engine Called SearchGPT
