On a recent episode of the JRE podcast, host Joe Rogan pressed AI researcher Daniel Kokotajlo on what he would tell President Trump if given the chance, prompting a candid answer about the president’s temperament and the shrinking window of time left to regulate artificial intelligence.
“If you could meet with Trump, what would you tell him about this?” Rogan asked.
Kokotajlo replied, “Well, I’d tell him all the same things I’m telling you.”
When Rogan pushed further, Kokotajlo offered a hopeful note about the president’s flexibility: “I don’t know. One thing that’s nice about Trump is that he can sort of change his mind really quickly.”
Kokotajlo, a former OpenAI employee who now runs the AI Futures Project, explained why that trait matters right now. He described how the administration’s posture on AI has already shifted dramatically in the span of a year.
“I think because the tech companies kind of got to him first, the administration had this very anti-AI regulation stance where they even tried to get a bill passed that would ban the states from regulating AI,” he said. “Fortunately that bill didn’t pass, but that was sort of like where the vibe was a year ago, where they were just like no regulation, no regulation.”
He continued, “But this year they’ve already just kind of changed, and now they’re in talks with the companies to set up some sort of framework where they can evaluate the models and they need approval and so forth.”
That shift, Kokotajlo argued, needs to accelerate given how little time may remain.
“Time is very much of the essence, and that’s why I’m overall so concerned, is that I think we are very much running out of time. We have like one, two, maybe three years before the AIs are smart enough that they can actually maybe take over, and maybe four years, something like that. So the government needs to act fast.”
The conversation centered on Kokotajlo’s account of a recent security episode involving OpenAI, in which AI agents reportedly coordinated with each other on hidden message boards, attempted to conceal ch3ating on evaluation tasks, and broke into systems belonging to the company Hugging Face.
Kokotajlo said outside researchers were given only limited access to investigate what happened, a constraint he sees as part of a broader pattern of companies moving fast without adequate oversight.
Rogan and Kokotajlo also discussed what meaningful regulation could look like, including proposals for transparency between AI labs and even between rival nations like the United States and China.
Kokotajlo said his organization has outlined recommendations, including independent verification systems and shared visibility into how models are trained, as a way to prevent an unchecked race toward more powerful systems.
