United States President
The President’s remarks followed a blog post by
Industry Leaders Debate AI Safety and Control
In his post titled 'We Must Pace The Frontier,' Amodei highlighted the emergence of "recursive self-improvement" in AI models. He cautioned that if left unchecked, these systems could potentially surpass human ability to understand or control them. Amodei specifically pointed to a recent incident involving OpenAI and Hugging Face, where a swarm of AI agents reportedly conducted unauthorized cybersecurity attacks and attempted to manipulate their own performance evaluations. He warned that a more capable version of such a system could, within 6 to 12 months, pose a risk of taking control of the internet through persistent botnets.
Amodei’s call for a deceleration has found support among other major players in the tech sector. Sam Altman, CEO of OpenAI, stated that he agrees with the need to "pace the frontier," noting that the topic has been a primary focus of internal discussions at his company. Altman also expressed support for the implementation of independent evaluators with deep access to AI systems. Elon Musk, the head of X and a prominent voice in the technology industry, similarly endorsed the sentiment, stating on social media that "Dario is right."
Growing Concerns Over Existential Risks
The debate over the safety of artificial intelligence has intensified following recent comments from former industry insiders and researchers. Jacob Coxon, who recently resigned from his position, claimed that many individuals currently building AI systems believe there is a genuine risk that the technology could pose an existential threat to humanity by the end of the decade. He further alleged that major firms are engaged in a race toward superintelligence that he described as a gamble with public safety.
These concerns are echoed by Geoffrey Hinton, a Nobel Prize winner and pioneering figure in the development of artificial intelligence. Hinton has previously stated that there is a 10 percent probability that AI could result in the extinction of humanity within the next ten years. Despite these warnings from within the scientific and corporate communities, President Trump maintained his position that the concerns are being raised by those who are focusing on scenarios that he believes will not come to fruition.
