You can, but very few people do.

Mostly what I see people doing is saying "Hey, Spicy Autocomplete, steal me some content from someone else to do <this thing>" and then post it up and ask why it doesn't work properly.

But then you've got things like the AI-based animation tools that Corridor Digital used to animate 3D scans of some of the crew's children's toys, to make a Toy Story-like video. It took damn near as long to do as it would have done by hand, because it all still had to be mo-capped and so on, but the results were incredible.

I guess it's similar to how just about anyone can pick up a cheapy flux-core welder and run a seagull-shit bead down a bit of metal, but a very skilled welder could do absolutely pristine work with the same crappy machine.

However, I still don't think it's a good idea to plonk small children down in front of a screen, and have a creepy AI voice read nonsense out to them. While I love the idea of working with the hallucinations of a dreaming computer, because I first read Neuromancer when I was 12 or so and grew up watching things like Blake's 7 with not one but two superintelligent talking computers, the reality is actually not really that good.

Children need people, not screens.