AI Illiteracy

AI Literacy Screenshot

Screenshot of a post by John Millionaire 2.0

 

 

 

The photo on Facebook shows an older gentleman asleep with a cat on his chest, both smiling in their slumber. The caption explains that the man is the poster’s father, and that this is the father’s “last picture and just that evening, he was no more and took away a part of us with him.

“You might not after all it [all stet], but this is also the last photo of [the cat]; in fact, just two nights after this photograph he also left us. There must be something about this, that Robin who was very devoted to my father, decided to go along with him as he couldn’t bear to be away from him. It is now that we picture them that way, snuggled in the way and in the arms that nothing and no one will ever take.”

The post has more than 20,000 likes, loves, and care emojis, and 772 shares. Scrolling through visitors, I see many are probably bots, but surely thousands are genuine. Three are my own Facebook friends and any number mutual friends with me.

The post is, of course, an AI fake.

We might start with an intuition that could turn out to be wrong: neither man nor cat looks anything other than healthy, safe, plump, content, and well-rested. AI is making fewer images with the “14 fingers and a grotesque laugh” vision of human life it apparently learned from advertising. But this “photo” (the only photons are from the screen and have never bounced off an original object) is kitschy, sentimental, and looks airbrushed. Part of the cat’s fur looks as if a lazy illustrator has tried to avoid fine details.

The wording in the caption is awkward too. Above all, the post is followed by a list of a dozen hashtags, such as #Netflix, Jennifer Lopez, Megan Fox, Hailee Steinfeld, Gal Gadot (and Wonder Woman), and Kendall Jenner fans, in a bid for clicks or to get friends for fake accounts.

Despite all this, the post also got 1,200 comforting and compassionate comments:

“What a beautiful photo and memory you have. I’m glad they went so close together. They will be in heaven together.”

“So beautiful!!! Prayers of peace, comfort, strength as you begin your grief journey for both of them!”

“Worth framing.”

There were only six laughter emojis and three angry ones, and very few comments such as, “hmmm. I’ve seen this post before, but with a completely different man and cat….”

“This is Bull – the exact same story was put up with a completely different person and cat – find a more ethical way of getting your comments and likes instead of a lie!”

“What bullshit.”

The post plays on people’s emotions to get attention, but what is so wrong with that? It is that thousands of people believed the fake post and acted on it. It is often mentioned that 54 percent of Americans read under a sixth-grade level. If visual-textual AI literacy is also this bad, what (else) will Americans believe about, say, politics, based on social media posts with altered or fictional material? I think we know but pretend not to.

People used to show us their family photos, albums, and home movies. We might have squirmed or questioned their interpretations of events, but we could be pretty sure the photos were not fake. Consumer-available digital manipulation was a big and recent step, but we did not process that societally before AI began to flood the platforms.

Be “politely paranoid,” New Scientist suggests. It is a good suggestion, but few will do it consistently.

The “photo” at the top of this post was reposted in a Facebook group called, *touches ground and looks up grimly* Something Boomer happened here. (The title is irresistible.) The picture is very obviously an AI fake. Look at the weird nametag, the way he wears a fatigue jacket but no pants. Note the disrespectful litter of American flags around his wheelchair, the awkward language and glitchy handwriting on his sign, which cuts halfway through his legs, and the Dr. Strangelove computer console. And why are his boots unlaced?

“Happy Birthday Solider [sic],” the caption by the original poster, “John Millionaire 2.0” says. This is followed by a list of hashtagged celebrities, as in the other post.

This post got 461,000 likes and loves, and a tidal wave of “patriotic” comments.

“All this [sic] people commenting have the right to vote,” one commenter said. “Fear the future.”

John Griswold

John Griswold is a staff writer at The Common Reader. His most recent book is a collection of essays, The Age of Clear Profit: Essays on Home and the Narrow Road (UGA Press 2022). His previous collection was Pirates You Don’t Know, and Other Adventures in the Examined Life. He has also published a novel, A Democracy of Ghosts, and a narrative nonfiction book, Herrin: The Brief History of an Infamous American City. He was the founding Series Editor of Crux, a literary nonfiction book series at University of Georgia Press. His work has been included and listed as notable in Best American anthologies.

Comments Closed