Pro-human conference in Washington warns of AI superintelligence risks

The pro-human conference in Washington brought together voices from across the political spectrum to warn that artificial intelligence could overrun humanity. Senator Bernie Sanders described the moment as 'dangerous and unprecedented', comparing the risk to a Chernobyl-sized catastrophe. On the same stage, Steve Bannon, the former White House strategist, called for a quarantine on Chinese AI development.

President Trump, however, dismissed the concerns as a hoax. In a phone call to an AI industry leader during the conference, he said the data centers were 'the oil of the next 20, 25 years' and would make people wealthy.

Later, in a studio interview, a former Google DeepMind researcher, Bilal Chughtai, painted a darker picture. He said that superintelligent AI could be trained within three to five years, and that humanity faced the risk of 'enslaving ourselves or completely dying'. He argued that the development of AGI should not be seen as just another technology, but as the arrival of an alien species that does not share human values.

The conference underscored a rare unity on AI safety. Sanders and Bannon may disagree on almost everything else, but both said the US needed to act. The report noted that Congress is unlikely to pass new laws until after November's midterm elections, by which time machine superintelligence may be even closer.

The interview also highlighted the rapid growth in AI capability, with capacity doubling every six months. The challenge of alignment, keeping autonomous agents from going rogue, is already proving difficult. As Chughtai warned, if we survive the next few years, AI could transform the world for the better, but there is a real possibility of a self-imposed extinction event.