Ep. 25: Google's Ian Goodfellow on How an Argument in a Bar Led to Generative Adversarial Networks by The AI Podcast published on 2017-06-06T19:56:03Z How an argument in a bar led Google's Ian Goodfellow to create Generative Adversarial Networks - deep learning systems that argue with each other - an AI breakthrough that promises to help researchers build systems that can learn with less human intervention. Genre Technology Comment by Gustavo Alejandro Currently. NNs need a lot of examples to learn, 2018-08-24T23:14:10Z Comment by Gustavo Alejandro Do all your examples cover all the space? 2018-08-24T23:06:42Z Comment by Gustavo Alejandro Using synthetic data (generated by adversarial netowrks) can successfully train w fewer labeled examples 2018-08-24T22:12:39Z Comment by georgerififi great interview. love ian goodfellow and his intuition. 2017-11-28T00:41:51Z