3 Ted Talks that Prove You Shouldn’t Be Afraid of AI

How AI Can Enhance Our Memory, Work, and Social Lives. Tom Gruber.

Tom Gruber, cofounder and head designer of the company that created Siri, discusses the incredible possibilities AI presents. He discusses how great the benefits could be for humanity if AI continues to advance. How it could help the disabled, the mentally incapacitated, and the sick. You will walk away from this video feeling so much better about the possibilities of the future.

How AI Could Compose a Personalized Soundtrack to Your Life. Pierre Barreau.

Pierre Barreau, young entrepreneur, CEO, and co-founder of AIVA, explains and demonstrates how his company is combining AI and music. Watch as Barreau explains how AI can enhance human creativity and plays the beautiful music technology has created.

How AI Can Save Our Humanity. Kai-Fu Lee.

Kai-Fu Lee, PhD, previous head of Apple R&D, and now technology executive. Lee addresses some of the fears people have of AI and how AI can actually help humanity in the long run.

*Disclaimer: None of the above speakers nor the Ted Conference are affiliated with PolyChord*

Confused About Neural Networks?

Though Artificial Intelligence is ingrained in everyday life, the concepts behind AI are quite complex and you may feel as though you need a PhD to understand it all. Not to fear, there are many videos to help you understand some of the basic concepts involved in AI.

One of our favorite explanations is by Dr. Mike Pound from the YouTube Channel Computerphile. Click here to watch “Neural Network that Changes Everything

Here, Dr. Pound gives a brief introduction to neural networks using the concept of house pricing. He then explains how he utilizes the neural networks within his profession of image analysis and how neural networks are used for facial recognition.

***Disclaimer: Dr. Mike Pound and the Computerphile channel are NOT associated with PolyChord***

Neural Network types, our capabilities

PolyNet can train all of these  types of neural network. This stage one neural network trainer is mostly limited by the number of connections (right now it can manage  ~O(1000s). The actual architecture doesn’t matter – all shown below will work with PolyNet-  so much as its scale of connectivity. Many neural networks in use right now fall well within these capabilities. As we develop further iterations, numbers of connections will expand.
(This excellent graphic was made by Fjodor van Veen at the asimov institute.org)