The internet is buzzing following Google's cloud roadmap announcements yesterday, along with the release of several AI web experiments which aim to help people understand neural networks by playing with them.
"Quick, Draw!" is a simple but highly addictive drawing game. The natural skepticist in me quickly wore off within a matter of seconds as I realised what the narrative bot voice was doing. As I started drawing on the white canvas, the neural network was analysing in real time what it could see based on every squiggle I was making. Not only this, but my results were then contributing to "it's" training, along with possibly hundreds of thousands of submissions from around the world.
I'm excited by Google's initiatives to help people understand these technologies better, as we often see "AI" commonly used as a marketing term to the general public, however my guess is that its not necessarily always well understood what the introduction of more advanced AI technologies potentially means for users.
I'd love to see this do what Pokemon did for AR (which is to bring awareness to the masses, even at if it was very rudimentary AR), but if it doesn't at least its definitely a great first step in the right direction.
If you haven't already, I urge you to give it a try here: https://quickdraw.withgoogle.com/
Word of warning: I take no responsibility for any productivity lost in your workday.
Quick, Draw! is a great way to familiarize yourself with how neural networks work to identify objects and text in photos, which is one of the most common forms of AI-guided software techniques we see daily on platform’s like Facebook and Google Photos. As you start to craft the doodle, Quick, Draw!’s software automaton will start yelling out words and phrases it thinks you’re trying to illustrate. As you get closer to the finished product, the voice starts to become a good indication of how your drawing could be misinterpreted as something else. If you’re on point, however, the neural network will hone in on the object and guess correctly.