Machine Learning
GAN
Two networks compete: one makes fakes, the other tries to catch them.
In real life
Some realistic fake faces online were made by two AIs competing: one makes fakes, the other tries to catch them.
Try it (2 minutes)
Play "forger and detective": one person draws a fake signature, another tries to spot it. Both get better each round.
Talk about it
Why might two competing AIs both get better?
In the game
Adversarial: when it's Audited, it hits back with 1 extra Flag. It costs 2 Data to play, and you can put it in your Training Set. It has Capability 3 (how hard it hits) and Trust 1 (how many Flags it can take), and scores 2 Tokens each time it Runs a Task. How to play