How a children's book gave me the creeps

Tim Harrower, a longtime friend, has just published a "children's book for adults," and I have ordered a copy based on some sample pages I saw on his Facebook page plus his past excellence as a writer and designer. He shared just two hours later in an email that things were a bit slow this winter and he wanted to experiment with AI tools. 

The book's text is all Tim, but the illustrations are all generative AI art tools (and I don't know which he used, but there are many available). The book is "The Dinosaur Owner's Manual," based on the concept that paleontologist Huxley Poot discovered live dinosaurs on a remote Pacific island in 1925 and brought them to Australia, where they flourished as pets for some time before adults banished them. This book is a reprint, so to speak, edited by Tim. It's a fun concept. 

It includes lavish illustrations, all generated by AI as guided by Tim's written instructions. He wrote: "You type in what you want (say, a dinosaur drinking tea), and the AI robot creates a painting, a photo, a sketch, whatever." 

I am sure it was quite a bit more complex than that, but his explanation makes sense based on what I have been learning while monkeying with ChatGPT to aid college writers. The best results from the AI "robot" come from more and more precise and structured prompts. That is likely true of AI art generators.

He has even created a promo video found here in which the narrator's Australian voice was AI-generated. 

I don't yet have the book in my hands but will not be surprised to find that the illustrations are rich and detailed. I know the text will be clever and engaging (and all from Tim). The book design will be terrific since Tim has long been a national leader in design theory and practice. 

I do have to admit that managing to illustrate an entire children's book gives me a moment's pause, so to speak. I would certainly feel a chill were I an illustrator, sensing that my services will soon no longer be required. 

Right now, I can't imagine the AI matching Tim's wits but, honestly, I would be foolish to simply shrug off that threat. It's just a matter of time, as least if an article in today's New York Times is correct.

The author has long been reporting on AI and is quite sure that we are, at most, a couple years away from AI tech that surpasses human capabilities in many areas. He may not be right, but if it's not two years, it will certainly be within ten. 

Today, we agonize over tariffs and an administration bordering on insane. We can't pass a federal budget and no one has any idea how to respond to a president who disdains all the accepted virtues and traditions of the American experiment. The stock market is tanking, though it will likely recover since rich people are quite clever and will do what it takes to "protect their own."

But the real challenge for the country and the world is likely to be generative AI that matches human intelligence and doesn't require 401Ks or even a rational government to operate. 

I used to think that the robots were coming, but not in my lifetime. I think I was wrong.