Anthropic’s head of communications scoffs at OpenAI

• OpenAI has had a wild week.
• The company has announced governance changes. And then key figures like Mira Murati and Bob McGrew left.
• An anthropocentric executive took the opportunity to mock OpenAI in response.

Tech rivalries are always a hot topic on X, and on Friday, an Anthropic executive couldn’t help but dig into OpenAI.

The social media spat between the two companies follows a wild week for OpenAI.

On Wednesday, the company announced that it is overhauling its governance structure to become a for-profit benefit corporation, no longer under the control of the company’s nonprofit arm. It also said it would give its CEO, Sam Altman, equity in the company for the first time. The changes marked a complete shift from OpenAI’s early days, when it was a little-known nonprofit organization.

Its chief technology officer, Mira Murati, who briefly replaced Altman during his surprise retirement last year, then announced he was leaving. Key researchers Bob McGrew and Barret Zoph said they would be leaving soon after.

It’s unclear whether the departures were related to governance changes, but tensions between OpenAI’s executives over its approach to developing general artificial intelligence and its commitment to safety, a core part of the nonprofit’s mission, have spilled over. repeated in public view.

OpenAI co-founder Wojciech Zaremba took to X on Friday to wax poetic about the latest high-profile departures. He compared the exits of Murati, Zoph and McGrew to children dying in the Middle Ages.

“Their departures made me think of the hardships parents faced in the Middle Ages, when 6 out of 8 children would die prematurely,” he wrote. “Despite the profound loss, the parents had to accept it and find deep joy and satisfaction in the 2 who survived.”

Zaremba said he has “cool memories” with each of the departing employees. With Zoph, he had a “fierce conflict about computation” that served as the genesis for OpenAI’s latest AI model, o1. He also said he had a fond memory of McGrew, who he said reprimanded him for “doing a jacuzzi with a colleague”. He did not elaborate. Murati, meanwhile, witnessed his engagement, he wrote.

Sasha de Marigny, a director at OpenAI rival Anthropic, was unimpressed by Zaremba’s historical reminiscences and allusions. Just a few hours later, she wrote on X: “Happy to report that the Anthropic co-founders are still happily working at the company. None were lost to the plagues of the Middle Ages or jacuzzis.”

Anthropic, by the way, is home to a number of former OpenAI employees, including one of its co-founders, John Schulman, who left to join a team at Anthropic focused on the safe development of AI technology.