What keeps Sam Altman up at night? OpenAI CEO reveals his darkest fears about AI

“When OpenAI first started, every article about us used the same Terminator photo because that was the only way to think about what AI was. If you asked people then about AI, they would say, oh yeah, it’s going to drive my car someday. Or like, oh yeah, that’s like the robots are going to fight us all – I’ve seen that movie. And that was really the way people thought about AI. And that was on my mind and a lot of our minds,” Altman told the host John Hope Bryant.

Altman went on to say that initially, OpenAI embarked on a somewhat frightening robot project, which was aimed at creating video game-playing agents for complex environments. He admitted that in hindsight, this idea seemed quite daunting to him. He stated that back then he and his team were quite young to fully comprehend the consequences.

“Originally we had a project with robots – obviously a little scary – and we were working on agents that could play video games, with the thinking that as they got smarter and smarter they could work in more complex environments. Now, we look back at that and think that was a scary idea. We were sort of young and naive at the time,” Altman said.

The CEO subsequently shared that the concerns about AI was mostly fuelled by media depictions. They also influenced OpenAI’s initial and ongoing concerns since its inception. He added that these narratives continued to shape the company’s approach and sense of caution.

“But all of those thoughts about the ways this can go wrong – you don’t need much imagination, because we’ve grown up with that in the media and it is such a compelling story, so when we started OpenAI we worried about that a lot. And we still do,” Altman said.

Altman is among the CEOs from Silicon Valley who have been actively batting for transparency and regulations in AI research and developments. The CEO has on several occasions highlighted the merits and demerits of AI which is rapidly proliferating across the world.