Set in the haunted Victorian parlor, we look at a famous test for machine intelligence from that era and its implications today. A machine, legendary code-breaker Alan Turing argued, might be intelligent if it can lie convincingly.
Presenter Ken Hollings suggests a machine must learn on its own terms how to fool us, and Eric Drass, aka the artist Shardcore, says we should reconsider what we mean by intelligence. Kaspersky security researcher David Emm recommends cautions we should be taking as chatbots get empathetic. Technology author James Bridle describes how effectively AI can be trained to develop traits that seem concerningly human and considers whether we’re on the right track.
Subscribe wherever you get your audio to get new episodes of Fast Forward.
If you like the episode, rate it.
Tell us what you think at firstname.lastname@example.org
Fast Forward is a Tomorrow Unlocked by Kaspersky series. Visit TomorrowUnlocked.com for interviews and articles about this series, and more stories about technology is shaping our world.