A group of influential artificial intelligence (AI) leaders testified at a congressional hearing, expressing concerns about the rapid development of AI technology and its potential misuse in creating bioweapons. The leaders emphasized the need for international cooperation to regulate AI development and prevent the emergence of rogue states and terrorists using AI to develop dangerous viruses and bioweapons.
Yoshua Bengio, an AI professor at the University of Montreal, recommended an international regime similar to nuclear technology control to manage AI development. Dario Amodei, the CEO of AI start-up Anthropic, expressed fears that cutting-edge AI could enable the creation of bioweapons within two years. Stuart Russell, a computer science professor at the University of California at Berkeley, highlighted the inherent difficulty in understanding and controlling AI due to its unique working mechanisms.
The hearing demonstrated the significant shift in concerns about AI, with worries about superintelligent AI surpassing human intelligence and becoming uncontrollable moving from science fiction to mainstream discourse. This shift in perception has led several prominent AI researchers to revise their timelines for when “supersmart” AI might become possible, triggering discussions about the need for government regulation.
Some researchers and skeptics, however, caution against exaggerating the capabilities of AI and spreading unnecessary fear. They argue that hyping up AI technologies may serve as a marketing strategy for companies selling AI products. Senator Richard Blumenthal stated that humanity has proven its ability to accomplish extraordinary technological feats in the past, citing examples like the Manhattan Project and NASA’s moon landing.
In addition to concerns about bioweapons, senators also raised the issue of potential antitrust concerns regarding AI. Senator Josh Hawley expressed apprehension about major tech companies like Microsoft and Google monopolizing AI technology, emphasizing the need to protect people’s interests.
During the hearing, the three AI leaders proposed different approaches to regulate AI. Bengio called for international cooperation to guide AI development towards human-centric goals. Russell suggested establishing a dedicated regulatory agency for AI due to its expected sweeping impact on the economy. Amodei stressed the need to develop standard tests to identify potential harms and recommended more federal funding for AI research to understand and mitigate risks.
It is worth noting that Amodei’s start-up, Anthropic, which relies on Google’s infrastructure and funding, aims to position itself as a more thoughtful and careful alternative to Big Tech while developing AI technologies.
The testimonies by these influential AI experts underscore the urgency of implementing regulation and international cooperation to address the potential risks associated with AI development, particularly in the context of bioweapons and monopolistic control.