Neural Networks, Kipu?

3min Snip

00:00

Play full episode

Summary

Transcript

Episode notes

Neural networks are like thousands of layers. The way you train a new network through deep learning is to show lat o pictures. And if tis a picture of a dog and it says a dog, then you want to go back and tell the network your bbasic gave the right answer. I think we'll get on to some of those issues around the personal targeting. Ok? So a deep neuror network would be likeThousands of layers. A connections have weight. It's just flowing forward a internally ructures that propagate these weights from the photo of the dog or cat itself,. Then it flows all the way to the end. And the last note either says dog or cat.

Kai-Fu Lee is one of the world’s leading AI experts and a bestselling author. He founded Microsoft Asia’s research lab that has trained CTOs and AI heads at Baidu, Tencent, Alibaba and Huawei. As President of Google China he helped establish the company in the Chinese market. And now, as CEO of Sinovation Ventures, he is investing in China’s high-tech sector, giving him a unique perspective on how AI is set to change our world over the next 20 years. On September 22 Lee came to Intelligence Squared to explain how AI is at an inflection point and urged us to wake up to its radiant possibilities as well as to the existential threats it poses to life as we know it. In conversation with Kamal Ahmed, former Editorial Director of the BBC, he discussed his new work of ‘scientific fiction’, AI 2041, co-authored with the celebrated novelist Chen Qiufan. The book offers up eye-opening scenarios of our techno-future – from a teenage girl’s rebellion when AI gets in the way of romance to a rogue quantum computer scientist’s revenge plot that imperils the world. To buy his new book AI 2041 with the Intelligence Squared discount click here: https://www.primrosehillbooks.com/product/ai-2041-ten-visions-for-our-future-kai-fu-lee-chen-qiufan/

Support this show http://supporter.acast.com/intelligencesquared.

See acast.com/privacy for privacy and opt-out information.

Learn more about your ad choices. Visit podcastchoices.com/adchoices