Google is very much in the business of artificial intelligence and machine learning these days – and some of its AI applications are incredibly fun.
Take for example Quick, Draw! – a hidden gem of an online game where players have 20 seconds to draw an object, while Google’s neutral network (aka, it’s AI system) tries to guess what you’re drawing.
Think of it as Pictionary against a robot.
You have six drawings to complete, and at the end you can look back and see why Google made the guesses it did.
You can even check out some other drawings of the same object by other players.
This also gives you an insight into how the learning process works for the AI.
Quick, Draw! is one of a number of AI experiments Google has placed on the web, all with the idea of learning as users play them.
The firm’s best example of this was their AlphaGo system, built by DeepMind, which learnt how to play the ancient board game Go and proceeded to beat the best player in the world, Lee Sedol, 4-1 in a best of five series.