The Monet Trick

 You can probably find articles on the Internet about the "study" where subjects were shown two a "Monet" painting of waterlilies and asked if it was a real Monet or an AI-generated image. Most got it wrong; the "painting" really was one of Monet's work (although a photographed image, which makes a difference), but most said it was AI-generated. 

Of this, Spencer Klavan says: 

"the Monet 'hoax' only shows that people on the Internet have opinions without much bases. They don’t know a Monet enough to discern if it’s real or an AI copy. Edmund Burke, in his wonderfully Burke-y sort of way, put it beautifully when he wrote that what sets good critics apart is chiefly the keenness of vision they develop 'from a closer and longer attention to the object.' The sensibilities take time to train. . . . My weirdly optimistic take is that all these artificial imitations aren’t so much displacing reality as reminding people what they love about reality, what they prefer about it to even the most impressive digital copy. Though cultivating selective taste in a chosen subject area requires an investment of long and laborious hours, it’s also something nobody can take away and no machine can replace, because it lives in you. It’s the weight of a real painting in a real gallery that you can stand before and be absorbed by. The books you grew up reading. Your blind grandmother’s garden.

I agree, but I think something else was going on. We are trained, implicitly, to be suspicious rather than gullible and to convince ourselves it can’t be a Monet and it’s a trick.