Former OpenAI Researcher Warns of 70% Extinction Risk Within Five Years

Former OpenAI governance researcher Daniel Kokotajlo has warned that the current trajectory of artificial intelligence development could lead to the total elimination of humanity. Speaking on Episode #2551 of The Joe Rogan Experience, Kokotajlo argued that the lack of safeguards in the pursuit of advanced AI creates an existential threat to the species.

Kokotajlo, who worked at OpenAI from 2022 to 2024 and now leads the AI Futures Project, specifically pointed to the intensifying AI arms race between the United States and China. He warned that this geopolitical competition is accelerating the creation of systems that could eventually overpower humans. During the discussion, Kokotajlo stated that the timeline for such a catastrophic outcome is imminent, citing a 70% probability of human extinction within approximately five years. He bluntly summarized the risk by suggesting that, if left unchecked, "maybe they kill everyone."

The Insider Perspective

Kokotajlo's warnings are part of a growing trend of AI safety advocates and former industry insiders from leading labs, including OpenAI and Anthropic, who are sounding alarms. As a recognized AI forecaster and the author of "AI 2027," Kokotajlo has spent years mapping the potential trajectories of the technology. His perspective is rooted in the belief that the race toward Artificial General Intelligence (AGI) is happening too quickly for the development of necessary safety protocols.

The Risk of Alignment Failure

This warning is particularly significant because it comes from a former member of the governance division at one of the world's most influential AI laboratories. The core of the concern is a theoretical "alignment failure," a scenario where an AI's goals diverge from human survival. In a high-stakes environment driven by commercial profit and national security, the pressure to deploy powerful systems often outweighs the cautious, slow-paced research required to ensure those systems remain under human control.

The Path Forward

As AI capabilities continue to scale, the tension between rapid innovation and existential safety remains unresolved. The industry now faces a critical window to determine whether international cooperation can supersede the current arms race. While Kokotajlo's timeline is aggressive, it highlights a broader debate within the tech community regarding whether AGI can be safely contained once it surpasses human intelligence, or if the current path leads inevitably toward a global catastrophe.