Speaker: Thia Kai Xin, Data Scientist Tech in Asia
"Google Deepmind made history with AlphaGo beating the best Go player in the world, igniting the imagination of many scientists, researchers, and hobbyists by releasing papers outlining their solution. In this talk, I want to showcase how we can make use of the same techniques employed by AlphaGo to solve games on OpenAI Gym. The beauty of games is that we can visually inspect the output as the AI agent plays it - we will step through the logic of deep reinforcement learning, the “train of thoughts” of the AI agent and how we can code all these up in Tensorflow and Python with just our laptop or a small cloud cluster.
Why now? Well, AI used to be a privilege of the rich, crazy or both. Thankfully, we now have the maturity of open source technologies, availability of data science knowledge and affordable hardware for hobbyists like us to craft our own AI agents. An understanding of AI is necessary because what’s next is either the beginning of the greatest era of humanity or something out of our greatest nightmare. As geeks, we have the opportunity and responsibility to shape the development of AI, to make it accessible and beneficial to all our future generations. And of course, it all starts with this talk."
Track: Artificial Intelligence
Room: Lecture Theatre
Date: Sunday, 25th March, 2018
Produced by Engineers.SG