Machine Learning
Reinforcement Learning
A model learns by trial and error, guided by a reward signal instead of labeled answers.
In real life
An AI can learn to play a video game by trying moves over and over and getting points when it does well.
Try it (2 minutes)
Teach a friend to find a hidden object using only "warmer" and "colder." That's learning from rewards.
Talk about it
What's a game where trial-and-error is exactly how you'd learn to play well?
In the game
Whenever this System is Deprecated, gain 1 Token. It costs 4 Data to play, and you can put it in your Training Set. It has Capability 3 (how hard it hits) and Trust 2 (how many Flags it can take), and scores 3 Tokens each time it Runs a Task. How to play