Sam Altman returns to OpenAI days after his shocking removal
It has been a crazy past week for OpenAI that saw the company’s CEO get sacked and then get reinstated days later.
Altman is co-founder and CEO of OpenAI and will be remembered for revolutionizing generative AI with the launch of ChatGPT last year. Under his vision, the small AI startup has quickly grown to be an industry leader, attracting billionaire investments from backers like Microsoft.
However, on Friday last week, Altman was abruptly removed from his CEO position by OpenAI’s board. Little explanation was initially given besides nebulous claims Altman had not been “candid” in communications. Confusion, speculation, and anxiety ensued inside and outside company walls.
It soon emerged tensions had been simmering between opposing factions. Sources described a rift between Altman and original OpenAI board members over how fast to advance AI systems to market.
As CEO, Altman had prioritized swift commercialization of new technologies like ChatGPT to seek profits, while board leaders preached more patience regarding safety risks from rapid innovation.
To add to the drama, Microsoft swooped in and offered to hire Altman and top executive Greg Brockman for a new in-house AI division. Brockman resigned from the company soon after Altman’s in what threatened to become a mass exit of employees from the company.
Workers rebelled, threatening an employee walkout if Altman was not reinstated. With almost the entire staff ready to flee, OpenAI risked total collapse under interim CEO Emmett Shear.
Just 72 tense hours later, OpenAI succumbed to the pressure and announced that Altman would return as CEO under a reconstituted board. Chairman Bret Taylor is an exited Salesforce executive, while economist Larry Summers will also join.
And while Microsoft won’t have the benefit of having Altman on their team it’s still a win for the company. Having stood with Altman during the attempted coup, Microsft has won themselves an ally and they can leverage that to further guide OpenAI’s direction.
What remains to be seen is whether commercial priorities will again override calls to implement guardrails that prevent the negative implications of rapid AI implementation.