Two years ago, with artificial intelligence the hot new thing, I enlivened a slow February morning by feeding Gemini AI this prompt: “Write a column defending the use of artificial intelligence in the style of Neil Steinberg of the Chicago Sun-Times.”
Gemini spat out a snappy patter that began, “Hey, Windy City. Steinberg here, with a word about our new neighbors: the robots. Yeah, I’m talking AI. Artificial intelligence. All the sci-fi buzzwords that used to belong to Isaac Asimov and now clog your Twitter feed.”
“Does that sound like my voice?” I asked readers, sincerely aghast. “Kill me now.”
Last February, I checked up on how AI had advanced by putting the same request to Gemini 2.0. The result was smoother, better, and — what really scared me — it provided a spot-on critique of what my writing is like unbidden, noting my “casual and conversational tone” and “cynical and skeptical outlook” seasoned with “A touch of self-depreciation.”
Who needs self-depreciation anymore? Not when reality depreciates us so regularly, with technology and the entire culture joining forces in a feverish drive to utterly moot what real people, such as myself, do for a living.
Now February has come again and is fast waning. So if I want to keep the tradition alive — the way things are going, there might not be another February — time to gird my loins and approach AI, bigger than ever, filling actual, real world jobs, so — cue the self-depreciation: I thought I’d lob Gemini 3.0 the same slow pitch and see what happens.
Four seconds after I clicked the button, it offered a solid headline: “The Ghost in the Machine is Just Us” and a column beginning:
“I stepped onto the Red Line this morning and watched a young man staring into his phone, his thumb flicking upward with the rhythmic precision of a piston. He wasn’t looking at the scenery — which, let’s be honest, is mostly the backside of crumbling brick three-flats anyway — he was arguing with a computer. Or rather, he was asking a computer to write a poem for his girlfriend because he, like most of us when faced with a blank page and the weight of human expectation, had found himself utterly empty.”
Sigh.
OK folks, using only that big, slow, imperfect lump of pink wetware between your ears, can anyone please tell me the problem with that paragraph? Not a subtle drawback either. I hope. Honestly, the flaw should reach out of the page and slap you hard across the face. Because — not to pressure you — if it doesn’t, it truly might be time for me to go out to Lake Michigan, settle myself on an ice floe and wait for spring, to be carried out to that Great Newsroom in the Sky, to share a lager with H.L. Mencken and Sydney J. Harris …
Anybody?
It’s not true. I didn’t step out onto the Red Line — indeed, would never use that phrase if I did, but name the specific station — there was no young man, no phone, no nascent poem.
Honestly, Gemini could have responded the same if I asked it to pinpoint the biggest problem with AI: firehosing a constant stream of fictitious slop.
This year’s Fauxberg column continues:
“The scolds will tell you this is the end. They’ll say that Artificial Intelligence is the final nail in the coffin of the human spirit, a cold, silicon reaper coming for our jobs, our art, and our very ability to think…. We are a species that delights in fearing our own shadows, especially when those shadows are cast by a new kind of light.”
That is true. Every new technology cues the funeral rites, back to Robert Louis Stevenson explaining how gaslight had ruined the romance of London. Don’t install the telephone in the nursery — children could catch infectious diseases over the wires! Radio would cause cows to stop giving milk and birds to fall out the sky dead. Television would erode children’s minds. Cell phones would give us brain cancer. AI will destroy mankind.
To me the key question is this: Will we continue to care if things are true anymore? Must the news have actually happened? Or can it be a scary, or funny or diverting story some machine tells to lull us to sleep? Are those Chinese robots really doing complicated kung fu moves?
We didn’t need AI to undermine the value of truth. Look at who we picked to run the country. Twice.
We see technology changing; what we don’t always see is ourselves changing along with it.
I could only use part of the quite good column Gemini 3.0 wrote. But I saved it for my files. After all, I’d written the thing, sort of.
