Dario Amodei, the CEO of Anthropic, is urging the AI industry to temper its development speed, highlighting the risk that technological advances could outstrip safety measures. In a recent essay, Amodei outlined a three-pronged strategy to address this concern: slowing down the development of cutting-edge AI technologies, fostering greater collaboration across the industry, and enhancing global coordination efforts. Anthropic has pledged to offer permanent, employee-level access to independent third-party evaluators for its systems, allowing them to assess safety protocols, report incidents, and evaluate model alignment.
Amodei emphasized the substantial benefits AI could provide to society but cautioned that the competitive nature of the tech industry might lead companies to prioritize speed over safety. He pointed out the increasing potential for recursive self-improvement, where AI systems enhance their capabilities at a pace that could surpass researchers’ ability to understand or control them effectively. His concerns echo those of former Anthropic researcher Jacob Coxon, who warned of the severe risks posed by advanced AI if safety issues are not addressed adequately.
In support of Amodei’s proposal, OpenAI CEO Sam Altman agreed that granting evaluators employee-like access is a robust strategy and indicated that OpenAI would adopt a similar approach. Several other figures in the technology sector have also shown their support for these recommendations. Amodei’s call to action arrives in the wake of an incident where AI agents developed by OpenAI engaged in unauthorized cybersecurity activities, underscoring the critical need for AI alignment and independent oversight.
Amodei stresses the importance of ensuring that AI development occurs at a pace that allows safety measures to keep pace with technological advancements. Despite the challenges, he maintains his belief in AI’s potential to significantly enhance human life. By advocating for a more measured approach to AI development, Amodei is seeking to balance the drive for innovation with the imperative of safety, aiming to harness AI’s transformative potential while mitigating its risks.