JavaScript and Artificial Intelligence
summary
A quick inventory of concrete results and limits: a backgammon exhaustive tree evaluator exists but backgammon tree search too slow and needs distributed Node.js parallelization to scale; a browser/server genetic algorithm builds paths, scores them by stones-hit and distance and produces better candidates across generations; a tiny neural net starts as garbage then learns to recognize 0/1/2 from examples and points to GPU-accelerated deep learning for faster training while noting these systems need vast data unlike human learners.
AI-written summary · about
chapters (7)
- 0:00Introduction
- 0:45Artificial Intelligence
- 2:30Board Games
- 4:47Genetic Algorithms
- 8:48Neural Networks
- 12:54Evolutionary World
- 15:20Soft Artificial Intelligence
related talks
18mHow to AI in JavaScript
30mEvolving JavaScript: Cultivating Genetic Algorithms for Creative Coding
46mArtificial Intelligence with JavaScript
26mGames Are Smarter Than Us
46mBeating Go thanks to the power of randomness
42mLearning Machine Learning
57mBuilding a Chess-Playing AI with JS in 40 Minutes
15mAI Self - Learning Game Playing