In June, I had a conversation with Ilya Sutskever, OpenAI’s chief scientist, at the company’s headquarters, which I reported in WIRED’s October cover story. One of the topics we discussed was the unique structure of the organization.
OpenAI started as a nonprofit research lab with the goal of developing artificial general intelligence (AGI) in a safe manner. The company found promise in large language models but required significant computing infrastructure and funding. Consequently, OpenAI established a commercial entity to attract outside investors, with Microsoft being a major partner. The majority of the company’s workforce transitioned to this new for-profit arm, which was restricted in its commercial pursuits. The original nonprofit’s board governed the entire operation and was committed to the initial mission.
Sutskever defended the company’s capped profit structure by emphasizing the need to ensure responsible development of AI, given its potential impact on jobs and society. The board was tasked with overseeing the commercial aspect to prevent the AI from getting out of control.
However, the same board unexpectedly dismissed CEO Sam Altman, citing communication issues that hindered its oversight responsibilities. This decision, made without advance notice to key stakeholders, sparked controversy and turmoil within the company.
Following insights into the board’s perspective, it seems they acted to uphold the company’s AI safety mission, prioritizing it over profit and employee relations. The dispute raised questions about the board’s ability to protect and monitor the mission without trusting the CEO.
While it remains uncertain whether Altman’s actions jeopardized OpenAI’s mission, the board’s handling of the situation risked damaging the mission. The repercussions were immediate, with Altman gaining more support and the company facing backlash.
A subsequent open letter from the majority of OpenAI employees condemned the board’s capability to oversee the company and demanded Altman’s reinstatement. The letter threatened mass resignations and a potential shift to a new AI research division at Microsoft, led by Altman and Brockman.
Initially resistant, the board later faced internal dissent, including from Sutskever, who expressed regret and reconciled with Altman, as evidenced by their recent interactions on a social media platform.