Can we trust artificial intelligence lie detectors? – Science Weekly podcast
by Presented by Graihagh Jackson and Jordan Erica Web from on (#43H95)
Liar liar, pants on fire? In this collaboration between the Guardian's Science Weekly and Chips with Everything podcasts, we explore whether it will ever be possible to build intelligent machines to detect porky pies
How good are you at lying? Could you fool a friend? How about a machine? We've recently learned that the EU is about to start trailing an artificially intelligent machine, or as the bloc calls it "deception detection", which is supposed to be able to detect if someone is lying at border control.
This got us wondering about lying - and about how we learn to lie. Could an intelligent machine pick up on subtle clues and cues, and if so, can we trust its judgment?
Continue reading...