Winning Card Games with 1000+ CPUs

Explore a fascinating journey into algorithmic game-solving and cloud computing in this EuroPython 2018 conference talk. Learn how Vincent Warmerdam and Bas Harenslak leveraged AWS Lambda and 1000+ CPUs to create a bot capable of mastering the card game SushiGO. Dive into the process of translating real-life game rules into algorithms, understand the need for extensive computational power, and discover why AWS Lambda is well-suited for simulation use-cases. Gain insights into implementing concurrency in Python, rapid Lambda deployment using Chalice, and the results of their experimentation. Examine the pros and cons of using AWS Lambda for grid search and simulation tasks, and uncover valuable lessons in both engineering and scientific approaches to problem-solving.

Intro
The problem
Optimization
Lambda
Jealous
Code Team
Asynchronous Processing
Lambda API Gateway
Cost of Lambda
Bruteforce
Converge
lambda user interface
lambda speed
price vs performance
engineering conclusions
scientific conclusions
play button
one more time
the truth
conclusion