I want to add videos to each my podcast episodes, so that I can post them on YouTube. One option I’m exploring is trying to sketchnote them while listening to them. This is a lot harder than I thought.

I have two problems. The first one is that I can’t do it fast enough. That’s easy to solve: I can just do it at any pace and fix it in my editor. The second is that I don’t draw well. Slowing down helps, but I just need to get better at doodling and coming up with objects to draw that represent the concept I am talking about in the podcast.

To practice, I saw Quick, Draw! on the Verbal to Visual YouTube channel. Google is training a neural net to interpret doodles, and it’s already good enough for this simple game. It gives you an object to draw, and you have 20 seconds to draw something the net recognizes as that object. When you are done, if it didn’t get your drawing, you can explore other drawings that the neural net did recognize as that object. They even let you explore the dataset of simple doodles. If you see one you like, you can tap it to see it being drawn, stroke by stroke.

I try to do it without thinking first because I want to get better at doing it live. I’m especially bad at animals. One drawback is that every prompt is an object, but I want to get better at translating verbs and concepts to visuals. I’d also love to see what people would draw for words like confusion, hunger, defiant, or tired.

Even playing for a half hour, I got much better. Not at animals, though.