Play blokus with computer, which will be smarter as more work put into it
each color plays purely random
use Monte Carlo Tree search to find the best move every time
generate large amount of game play data and train classfier (Neural network, etc.) on good / bad move
reinforcement learning --> smarter AI in while playing (memory-based recurrent neural network)
algorithmic game theory based AI --> could form team / foe with different colors, collaborating or sabataging collaboration of AI agents
deploy game on server using Heroku
do some other projects (such as flying Alexa networks with emotion detection [already purchased hardware])
AI agent - DONE
human player interface - NOT STARTED
Refactor code for future dev - NOT STARTED
g++ -std=c++11 -Wall -Werror -pedantic -Wextra -Wconversion -O3 main.cpp grid.cpp shape.cpp player.cpp game.cpp -o blokus && ./blokus