AI lets your fingers do the animating

I got a kick (heh, oh no) out of this fun marriage between organic & imagined worlds. If only I’d had this tech in 8th grade, I’d have used it on my buddies’ compulsive finger-scale simulation of kick flips & other skateboard tricks.

The animator’s comments are pretty thoughtful, so in case Twitter won’t display its translation below, I’ve pasted it below the post:

Ever since I started dabbling in animation—and it’ll be four decades soon—I’ve been waiting for tools that let you animate in a more organic and contextual way.

The entire traditional technical overlay on computer animation is absurdly overcomplicated.

It worked a bit like a gatekeeper. “Look how terrifyingly difficult this is, better leave it to the specialists, because no one else will grasp the depth and complexity of this magical process.” In fact, there are few other industries that have turned the complication of tools into a virtue, while simultaneously making the process inhuman and unintuitive.

From the start of this fun, I’ve dreamed of something else. Building the dynamics of a scene or a character’s movement is hard on abstract Bézier curves, on multiply nested hierarchical skeletons, but it’s easy to show with a gesture, fingers, body, or even just a whistle.

Finally, it’s possible. People are multimodal, so it’s great that machines finally are too. Finally, instead of “move the mouth corner controller 5 mm up, and increase the eye squint, move the eyebrow controller up 3 mm”—you’ll be able to say “smile.” Which is what I wish for myself and for all of us.