Sam Altman Becomes Mark Zuckerberg
Last modified
When it comes to new technologies, Sam Altman is becoming more like Mark Zuckerberg.
OpenAI CEO Sam Altman is increasingly becoming like Facebook's Mark Zuckerberg, pursuing a "moving fast and breaking the mold" strategy while also avoiding PR disasters and issuing apologies.
Even as the potential dangers of AI loom large, Silicon Valley continues to champion leaders who take a “forgiveness rather than permission” approach.
OpenAI recently announced that it is establishing a new safety and security committee, while also starting training its next-generation large-scale language model, GPT-5. This is a typical Zuckerberg-style strategy: apologizing for a problematic incident with bureaucratic measures, while simultaneously rushing to the next “fast-track” project.
Sam Altman has apologized for a “communication breakdown” over the use of Scarlett Johansson’s voice in ChatGPT, and for the equity clawback clause in OpenAI employees’ severance agreements. Amid the controversy, there have been notable departures from OpenAI’s safety team.
In both cases, the apology didn’t stop the news cycle. OpenAI provided the Washington Post with documents defending its hiring of actors to give ChatGPT a female voice. And in an interview with Bloomberg, Altman acknowledged that he had signed a restrictive nondisclosure agreement that he claimed he didn’t know about, and he confirmed that he would release former employees from those agreements.
Zuckerberg has apologized for mistakes for almost the entire time Facebook has existed. Yet the company has pushed through policies and plans, and user numbers, engagement, and revenue have continued to grow. Similarly, despite recent setbacks, Altman’s OpenAI has been able to navigate controversy and quickly release dazzling new AI demos and capabilities.
Both CEOs have been criticized for being overly optimistic about the good potential of their companies and naïve about the potential harm they could do. But in the startup world, where recklessness is often rewarded, founders tend to “ask for forgiveness rather than permission” and “move fast and break the mold.”
But Altman appears to have learned from Zuckerberg’s mistakes. He has adopted a more candid style of communication and regularly addresses the inherent risks of OpenAI’s mission. In particular, in his first congressional testimony in May 2023, he recently appeared less robotic than Zuckerberg and more willing to work with lawmakers.
Some of Facebook’s biggest mistakes have resulted in real-world harm in the offline world. But mistakes in AI deployment can result in even bigger nightmares. We need to keep an eye on this.
Subscribe to '오늘배움'
Get the latest on education trends, AI-based learning know-how, and corporate talent development insights at Learn Today. Receive useful news as we explore the future of digital learning together.
