alphago

The following video talks about the speed of advancement of artificial intelligence. The video come from the ColdFusion YouYube channel which is quickly becoming a Dawn of Giants favorite. If you haven’t visited their channel, you show definitely stop by and browse some of there videos. The video primarily focuses on neural networks and how they are advancing more by means of self improvement (sounds like a self help program for AI… which, I guess in a sense, it is) rather than updates created by human programmers. The video gives an overview of neural nets and also presents several fascinating examples. AI cameos include AlphaGo, WaveNet, StackGan (a generative adversarial network), and more.

Runtime: 12:07

This video can also be found here. Make sure to stop by and give them a like.

Video Info:

Published on Apr 12, 2017

Was hacked, I’ll have to rewrite this later.
Sources:

https://deepmind.com/blog/wavenet-gen…

https://drive.google.com/file/d/0B3cx…

https://techcrunch.com/2017/03/07/her…

TED Talk https://www.youtube.com/watch?v=BfDQN…

Sergey Brin: https://www.youtube.com/watch?v=30rx3…
https://futurism.com/google-ai-secret…
https://qz.com/877721/the-ai-master-b…

https://www.wired.com/2017/01/mystery…

https://www.forbes.com/sites/aarontil…

https://futurism.com/new-ai-can-write…

//Soundtrack//

6:18 Cash – Hope

https://soundcloud.com/cash-official

Category
- Science & Technology
License
- Standard YouTube License

This video explain’s Google’s DeepMind AI. The goal of the DeepMind project is to develop general purpose algorithms which, along with the development of artificial intelligence, will teach us more about the human brain. Part of the DeepMind research was a project called AlphGo which beat Go world champion Lee Sedol in 2015. Due to the intuitive nature of the game, it was not expected that an AI would succeed in beating a Go champion for at least another decade. DeepMind uses a type of machine learning called deep reinforcement learning. Machine learning is not to be confused with expert systems. Expert systems work within predefined (pre-programmed) parameters whereas machine learning relies upon pattern recognition algorithms.

Runtime:13:44