I volunteer in a homeless shelter, and in the expressive therapy studio - where I hang out - we have light instrumental music playing in the background. (OK. Maybe it's muzak, but it makes for a pleasant, calming environment.) The music/muzak is accompanied by the cheesiest AI pictures and videos: flowers in colors never seen in nature; cozy little houses far more fairytale than anything that Thomas Kincaid ever dreamed up; quaint little town scenes. Fakery all!
We put one on the other day that purported to be a drone's-eye tour of Salem, Mass. My sister lives in Salem, and I've been there hundreds of time. I didn't recognize a thing. A bunch of generic New England-ish, with some witch stuff thrown in. Dreadful.
Another recent un-favorite was a scene of a French cafe in a rain storm. The windows were wide open and gouts of rain were pouring in. Yet there was no rain on the tables. There was a blazing fireplace in a location that no fireplace would have been located in a billion years. And if you're going to have food in a French restaurant sitting on a table, it should at least look edible. What was presented might have been found in a pub in Scotland, but not in France!
Then there are the fake YouTube videos that are popping up. And the fake pictures that are all over the web.
Most are terrible and obviously AI-generated. But AI is getting better and at some point it will be impossible to spot whether something is real or not.
Remember when we used to quote Chico Marx in Duck Soup, "Well, who you gonna believe? Me or your own eyes?" (Yep. I thought it was Groucho, but I looked it up and apparently it was Chico.)
Our own eyes are, increasingly, going to betray us.
But it seems that everybody's doing it, doing it, doing it.
Case in point: Stanford University.
A couple of years ago, their dining service posted a pic of three typical students getting ready to chow down on noodles in celebration of the Lunar New Year.
Nothing special about the kids. Pleasant, attractive students. An Asian woman; a couple of fellows, one somewhat Hispanic looking. That Hispanic-looking guy was Billy Ramirez.
Recently, Stanford decided to borrow the image and use it on banners. Only Billy Ramirez, in his GAP sweatshirt, is no longer holding his plate of noodles. No, a Black woman in a Stanford sweatshirt is. The appearance of the other two students has been marginally altered. The Asian woman's face was made thinner, and I'm not quite sure what they did to the kid in the middle. His hair is darker and his face appears rounder. In the original shot, he seems as if he might be Hispanic. Turns out he is. But they've made him seem a bit more generic. Is he or isn't he? Only his photographer knows for sure.
Mr. Ramirez, a senior who is majoring in human biology, said there was “a comedic element” to the matter. Nevertheless, he said that he was frustrated by Stanford’s move, which he learned about from a friend.One issue here is, of course, trying to convey diversity - which I believe is a good thing - where there is none.
Stanford acknowledged in a statement on Tuesday that it had used artificial intelligence to change the original image.
“Stanford’s A.I. policy is clear: The use of A.I. in producing or altering images of Stanford people, events, research, facilities or achievements is strictly prohibited,” the university said. “Both the alteration and lack of disclosure in this case violate that policy. We are working with Residential and Dining Enterprises to provide additional training and ensure proper review of all materials.” (Source: NY Times)
No comments:
Post a Comment