enter image description here
Machine learning, deep neural networks, computer vision, natural language processing, multiagent reinforcement learning, artificial “general” intelligence/superintelligence, you name it. The 8 day mini-course Artificial Intelligence: Principle and Practice will cover it.
We are going to peer far beyond mathematical symbols on paper into the intelligent system they define. We will learn a huge spectrum of deep learning approaches, techniques, and perspectives including classical artificial ‘intelligence’, machine learning, deep learning, reinforcement learning, and the generalization limits of artificial intelligence. The lab will give hands-on experience with real data, real experiments, and real learning experience using several deep learning frameworks and tools (scipy and IPython suites, jax, tensorflow, tensorboard, pytorch, huggingface, tf-lite, openai-gym, pettingzoo; TDW, generic docker deployment) as well as using cloud compute to manage Internet-scale deep learning workflows. By the final week, you will have (individually or in teams) designed and developed your own novel, real world AI system, and on the final day give a presentation on it.
Lectures run 1:00-1:45pm and labs 2:00-2:45pm M/W or T/Th. There will be no homework, no tests, and no grading. We’ll save the tests and benchmarks for our robots. This course is not accredited However you can be sure there will be machine learning and more importantly human learning. If you’re neurons are on the verge of firing and your reward estimator feels like it’s ready to explode, please signup at https://jacobfv.github.io/Artificial-Intelligence-Principle-and-Practice/.
ps: (Much of this document was drafted using artificial intelligence.)