Audio & Video

Fake news: you ain't seen nothing yet


Normally, computers generate speech by linking lots of short recorded speech fragments to create a sentence. By trying to fool the adversary, the generative software would learn to create images that look real, but are not. In the past five years, software powered by similar algorithms has reduced error rates in classifying photos from 25% to just a few percent. Mr Goodfellow now works for Google Brain, the search giant's in-house AI research division (he spoke to The Economist while at OpenAI, a non-profit research organisation).