Do you remember my post about Google AutoDraw? If not, go read it. Then come back.
Welcome to the next sentence. For those of you who don’t know, Quick, Draw! is a Google experiment where they ask us (humans) to doodle specific things in a short amount of time that are then collected into an open-source data set which is then fed to neural networks (not humans) by developers (humans) to train them (not humans) in patterns and how to identify objects based on what we (humans) have provided.
Phew. Still with me?
Since the data is available for anyone, it gives us as designers (and anyone, really) an unprecedented database from which to see how our fellow humans think. We can then take that information and do with it whatever we please. This is all incredible. What is not incredible, however, is how utterly unimaginative we (humans) are. While we enthusiastically threw ourselves to the task of doodling (50 million drawings!), we often went with the iconic idea of the thing to draw rather than drawing the actual thing itself.
When asked to draw a telephone, I drew my current phone – a flattish, rectangular block with rounded corners and little square app icons. Quick, Draw! did not understand what I was drawing. It kept suggesting “oven,” “box” or “calendar.” This was distressing, so of course I went to see what a telephone looked like according to Quick, Draw!
Hello, rotary phone! Do you remember the last time you saw or even touched a rotary phone? For you youngsters, have you ever done either of those things?
Apparently, pigs are only disembodied circles within circles. I dread the day when a Quick, Draw!-taught neural network sees a real pig.
Looking at what people have drawn for “elbow” left me swimming in a sea of open-ended boomerangs with the occasional line pointing to the pointy-bit. What was probably the most amazing one to browse through (for me) was The Great Wall of China! How do you draw The Great Wall of China in 20 seconds? I have a sneaking suspicion the team behind Quick, Draw! put that in just for their own enjoyment. The results are pretty consistently things that look like teeth, squiggles, or just black boxes of obvious frustration.
Passing the time perusing those 50 million drawings is both entertaining and instructive. As a designer and illustrator, I’m often investing large amounts of thought and energy into conveying a non-verbal message in the hopes that I won’t need to draw an arrow to the pointy bit or write “CHINA” over my squiggly lines meant to be The Great Wall of China (not that that ever helps). Seeing this large body of generalizations willingly contributed by thousands of individuals all over the world is pretty staggering. UFOs are little discs with a bubbly dome on top, alarm clocks are analog, lions are only males, flowers largely look like daisies, and the way people doodle eyes is terrifying.
This kind of drawing under pressure has been explored in other arenas of design as well recently. Back in 2009, an Italian artist named Gianluca Gimini doggedly pursued strangers and friends with the task of drawing a bicycle from memory. He then 3D printed the results. Please, I beg you, go look and have a belly laugh for me.