As is well known to my loyal reader(s), I've had a lot of fun making artificial intelligence jump through hoops to create various images of wemics. Well, no surprise, the state of the art continues to advance — today Google Labs invited me to try its new Whisk tool. It is intended to let you pick a style and then either use text or to upload an image, mix it into a scene, if you like, or just use their presets. And the presets are a hoot!
I started with this image, a pretty good liontaur I made with Night Cafe, a service I have trained on about fifty-odd wemic images I had lying around, mostly stolen by me from artists around the web. Boo hoo, cry me a river. I used your image to train an AI. You and the New York Times can sue me for all the money I've made doing that.
You can click on all the images in this screed to embiggen them, by the way.
So I fed this image into Google Whisk and told it to make me a plushie!
Then I used that image to make a toy in a bubble like you get out of a gumball machine.
Then one of those enamel pins.
Mister Gump says life is like a box of chocolates. I guess Google took that seriously.
But in the end, nothing beats a bento box.
Honestly, this tool leaves me a little speechless. I guess the next step is to try out some of the animation tools, and see how they work. A fun project for the future.
