Categories
Letter From The Editor Opinion

Imagining Intelligence

“Imagine a tater tot ice skating.” 

“Oops! I can’t generate that image. Can I help you imagine something else?”

“Imagine a tater tot on ice skates.”

“Oops! I can’t generate that image. Can I help you imagine something else?”

“Imagine a tater tot.” 

“Oops! I can’t generate that image. Can I help you imagine something else?”

“Imagine a cylindrical fried potato on ice skates.” 

The above is a “conversation” I had with Meta AI after several prompts appeared in my Facebook feed encouraging me to “Ask Meta AI to imagine anything.” Examples given included “imagine a clown in Paris,” “imagine an alien wedding,” “imagine running on Mars.” Whatever you choose to “imagine,” the AI feature will produce an image for you. Or not. But I had to give it a try. In thinking of something random and absurd (and while likely craving potatoes, which, if I’m honest, is often), I asked it to imagine a tater tot on ice skates. Psh! Turns out the thing doesn’t even know what a tater tot is. Intelligence, shmintelligence! Big dummy! I got a chuckle when it finally dreamed up my more specific request (result shown here) — spiralized fried potatoes atop the cutest little pair of skates. You’d almost think it’s real! Or not …

One majorly unfortunate issue with AI images, audio, and video is that there are a whole lot of folks out there who can’t tell it’s AI. In the last couple of weeks as hurricane coverage took over our social media feeds, loads of AI images appeared — of Donald Trump in jeans, waist-deep in floodwater; of crying children clutching their dogs in heavy rain. On one thread below a post of the former, commenters were split. The image was real, some demanded. “Bless him. He was there doing all he could to help.” Others agreed that, yes, maybe the image wasn’t authentic — completely “imagined” by AI, to be exact — but regardless, they said, “The sentiment is real.” The sentiment is real. Since when does sentiment equal reality? Some have bigger imaginations than others. 

Speaking of meta, I also asked the image generator to “imagine artificial intelligence imagining artificial intelligence.” The result was a shiny robot — a hodgepodge of metal plates in the shape of a human, with some exposed wiring and small lights glowing from within its head — hands rested on a keyboard, intently studying a computer screen with indecipherable lettering. Imagine you imagining yourself. Woah, things are getting heavy! 

Jokes aside, things are getting heavy. As we near the election, misinformation abounds, much of it generated with AI and spreading like wildfire online. Users are even making video clips of what appear to be Biden and Kamala saying things they’ve never said, or images of political figures “imagined” to be standing arm and arm with U.S. adversaries. And an astounding selection of American voters believe, apparently, everything they see. I wish I had that kind of imagination! 

Another recent example of rabid imagination fueled by AI, shared by a Facebook friend, was a video of “robots doing field work.” “So what do we do with human[s]? Wow!” he captioned his share of what was absolutely in no way real — cartoony even — showing one Short Circuit-looking robot, scythe in hand, gathering bundles of hay and another shoddily “imagined” robot holding onto a bull’s tail and standing (dancing?) on top of some sort of seeder or tiller pulled behind it. The whole thing looked straight out of a poorly programmed video game. But alas: “Scary. We will lose jobs,” one commenter noted. “Sad day,” another lamented. 

It is, in fact, a sad day when a percentage of the population cannot tell the difference between what’s real and what’s fake. Maybe robots will take our jobs. Maybe somewhere deep in his twisted heart, Trump imagined himself wading through waters and rubble helping hurricane victims. The sentiment is real, they say. Who cares if what we base our opinions on is complete and utter nonsense? 

I, for one, would like to imagine that people aren’t so gullible. That they’re capable of discerning truth from blatant lies. That somewhere out there, actual human intelligence still exists. One thing I know for sure, though, is that tater tots are real. Meta AI cannot convince me otherwise. I think I’ll go have some. Imagine that!