A.I. Is Getting Better Fast. Can You Tell What’s Real Now? – The New York Times

“A.I. Is Getting Better Fast. Can You Tell What’s Real Now?

“Artificial intelligence tools can create lifelike faces and realistic photographs — and they are getting better all the time. The phony images now appear regularly on social media, with many users seeming to believe that the images are real. But there are still some telltale signs that an image was made by A.I.

Can you tell the difference? Take our quiz.

1. Is this celebrity photoshoot real or A.I.?

A.I.
Not A.I.

Oops, not quite. This is a real image. This image shows some stars of the “Justice League” movies — Ben Affleck, Gal Gadot, Henry Cavill and Jason Mamoa. It’s a genuine image but it may look slightly unusual because it was either cropped or compressed after it was reposted several times on social media. The original image is below.

. . . . . .

David Lindsay Jr.
Hamden, CT  NYT Comment:

Reading the comments is instructive. The reporting is excellent. Thank you. At 60% I did much worse than expected, but this was my introducdtion into what to look for. I seems that the only logical thing is to ban all AI generated images that are not specifically for scienctific or military matters. All AI generated images shoud be required to be labelled cleary as AIG, possibly in all four corners and one or three in the middle, top, center, bottom. Penalites for using unlabeled AIG should be severe. The public should not have to doubt every photo it sees. Otherwise, won’t this be dozens of more nails into the coffin of democracies? InconvenientNews.net

DL: I hope every reader can access this article in the NYT, take the test, and read the comments. You shoud be allowed 10 articles at the NYT per month as a non-subscriber.

If you have no views left, try to put your browser into incognito mode to access this article.