As the field of artificial intelligence (AI) has grown from vague hopes to striking realities, as seen in technologies like ChatGPT, so too have alarms grown clearer. Geoffrey Hinton, the so-called “godfather of AI,” left Google to speak more clearly about the threat posed by the technology, which may one day grow smarter than its creators. He strictly opposes allowing the military to use AI and worries that an artificial hyperintelligence could one day manipulate human beings.
How do we convince our machines to behave ethically, even when we’re not watching? A forthcoming book by Eve Poole called Robot Souls: Programming in Humanity argues that we have to make them more like us, and that means imbuing them with empathy and compassion, even if it means reducing their efficiency. This would be no simple program update, and as with humans, empathy and caring would remain specific to the individual, a kind of artificial subjectivity.