Google’s AI chief warns of growing AGI risks (www.techradar.com)

Google DeepMind's CEO, Demis Hassabis, has issued a stark warning about the risks associated with the impending arrival of artificial general intelligence (AGI). In his recent essay, he acknowledges the potential of AGI as a transformative achievement for humanity but urges the need for urgent regulatory measures to address emerging threats. He highlights concerns about cybersecurity issues and the possible escalation of biological and nuclear risks as AGI capabilities advance, emphasizing that even AI developers are becoming increasingly anxious about the uncharted future of this technology. Hassabis proposes the establishment of a new US-led regulatory body, funded by the industry but federally governed, to evaluate AI systems before their release. This body would conduct thorough assessments focusing on cybersecurity and safety measures, and could even slow down releases if necessary. While he remains optimistic about AGI’s potential to drive breakthroughs in fields like medicine and energy, he stresses that societal discussions on its implications, such as wealth distribution and control over advanced systems, are critically lagging behind technological advancements. This moment presents a vital opportunity for society to shape the future of AI responsibly for the benefit of all.