Can Machines Come Alive?
Could intelligent machines challenge humanity's place on Earth? In recent years, this staple of science fiction has become a legitimate question. Many experts reject the possibility, but credible theorists argue for an upcoming 'singularity' that produces superintelligent artificial intelligence (AI). What happens next is debatable.
Some welcome superintelligence as a kind of angel. Others detail existential threats from hostile machines. One thing is certain: current rates of progress demand that we consider scenarios where humanity shares the world with a powerful new species. It is surprising that discussions of superintelligent AI neglect biology. This lecture discussed recent advances in systems biology, and applies them to the question of technological progress. Biology introduces a radical shift from intelligence to agency. As recent history shows, a virus can be deadlier than our smartest machines. The lecture proposes a new framework for evaluating the possibility of truly autonomous systems, and, if necessary, sensing and regulating threats from them.
About the Philosopher of Science
Ali Hossaini works at the cutting edge of art, technology and science. His artwork Ouroboros was acclaimed by the New York Times, which calls him “a biochemist turned philosopher turned television producer turned visual poet.” He is co-director of National Gallery X, one of the founding partners of UKRI Trustworthy Autonomous Systems Hub, and in 2020 he hosted two conferences on trustworthy AI. He recently contemplated possible futures of AI in three creative productions: The First, GROUPTHINK, and The AI Gallery. In a special issue of RUSI Journal, four biologists respond to his essay Modelling the Threat from AI: Putting Agency on the Agenda.
