What’s this? An actual innovation by the Republican National Committee? And it’s not winning the White House by a few marginal county victories in three states? Nor is it repackaging Reagan one-liners in a novel fashion, e.g. “Freedom fighting more than ever cause Constitution and liberty and tax cuts and those damn Ruskies at it again!”?
No, it’s actually something new? Well, I’ll be dipped. The party of permanency breaking the mold. And it’s not Caitlyn, née Bruce, Jenner giving an own-the-libs token talk at the Grand Old Party’s national convention.
You may not have noticed, but the current President of the United States is running—or rather gimping—for re-election at the spry age of fourscore. The announcement may have escaped your attention because President Biden phoned it in, with rotary dial, operator service, and Ma Bell. There was no rally, no packed crowd, no bunting, no tiny American flags, no-spittle-sprinkled impassioned pleas by party grandees, no event whatsoever. Team Biden dropped a b-roll-brimmed YouTube video with two, count-’em-two, seconds of the president actually speaking to the camera.
So how does the RNC combat this piddling show of electoral force? By being even lazier: releasing AI-generated video response, featuring Sleepy Joe and glass-gurgling Kamala Harris rendered in Nintendo GameCube graphics and an apocalyptic voiceover narrative straight out of the Branch Davidian compound.
What’s new about the GOP invoking Book of Revelations-esque civil unrest? Didn’t Nixon already pick that play with Lombardian success?
The difference: the shameless use of artificial intelligence. And RNC spokesperson confirms it’s the first time the historically hidebound org has produced a 100% AI hit job. According to Vice News, “it’s the first campaign video from a major party to do so in history.” Cameron Joseph, the reporter making the historic charge, should rewatch dim-flash-in-the-pan Senate candidate Christine O’Donnell infamous “I’m not a witch” spot, which, if it can’t be classified as artificially intelligent, definitely lacked any intelligence whatsoever.
Joseph smells trouble anyway, asserting “no one knows how to handle” AI-produced ads. He excitably postulates, “The time when a deepfake videos which can’t be easily discerned from real events play a major role in campaigns seems not only inevitable but closer than ever.”
He’s not the only one fretting the democratic mindscramble caused by algorithmically enhanced political messaging. “We’re not prepared for this,” warns VP of Intelligence A.J. Nash of ZeroFox, the “this” being full-on synthetic political adverts. Nash’s premise is, of course, ungrounded. It doesn’t matter how prepped we are—Skynet continues forth regardless of our wishes.
Oren Etzioni of the Allen Institute for AI gives us the following frightful scenario: “What if Elon Musk personally calls you and tells you to vote for a certain candidate?” First off, if Musk personally dialed me up, I’d give him an earful over making Twitter a dysfunctional mess. Then I’d thank him profusely for ruining the journo class’s favorite plaything. Only after that would I hear him out on why Tim Scott can heal our nation’s racial and moral divide, or whatever sentimental pap he’s peddling. Nevertheless, Etzioni answers his own question with “A lot of people would listen. But it’s not him.” The “a lot” is doing a lot of presumptive work. With a vague sum, Etzioni thinks everyone who’s ever gotten a direct-mail form letter signed by a sitting president literally thought the chief executive was taking a 20-minute break from bombing the Mideast to personally request $5. The American public isn’t the brightest, but it surely realizes by now that Ted Cruz isn’t actually typing out dozens of emails a day asking for a donation to send more patriots to the Senate.
Podcaster and winter-outwear model Tim Pool is also perturbed by AI ads, lambasting Ron DeSantis’s limping campaign for using generated smear images of rival Donald Trump. The dynamic has gotten so worrisome that Rep. Yvette Clarke has introduced a bill to require disclosure of any AI-produced parts in political ads, so that when Team Trump inevitably releases a XXX-clip of “Meatball” DeSantis and “hirsute” Biden caught celebrating Pride month in flagrante, Americans won’t be too scandalized.
All this worry, concern, fretting, nail-biting, hand-wringing, and end-times talk of electorally weaponized AI assumes both a lie and a truth: that politics heretofore has been a gentleman’s game based in verity and that the public believes everything it hears. It’s not an authorial conceit to admit that the average American voter is a few epistemological rungs below Publius when it comes to civic history—George W. Bush was elected twice after all. Gullible as the public may be, it, including you and me, responds to messaging that chimes with our political priors. Even if the Biden campaign puts out a deepfake video of Trump kicking his way through a field of puppies, at least half of the Democratic Party will believe its veracity. The obverse is also operable: if a Trump rival releases an attack featuring boxes of illegally obtained classified docs in the Mar-a-Lago powder room, a sizeable chunk of the MAGA base will claim the photo is fake news.
Why? Besides Mencken’s dictum that “democracy is the theory that the common people know what they want, and deserve to get it good and hard”? We readily believe our preferred candidate and party because the American demos has been thoroughly prepped for buying into seemingly real ads after decades of consuming false narration from our political-media complex. What’s a bigger threat to the body politic? A computerized image of Hillary Clinton sucking the rejuvenating blood of infants? Or deliberate lies like “Saddam has WMDs,” “you can keep your doctor under Obamacare,” and “Trump colluded with the Russians to win the presidency”?
The best programmed AI in the world could never come up with a fantastical story like the sitting president paid Muscovite minxes to pee on his predecessor’s bed. Hell, most fiction writers wouldn’t craft something so straightforwardly bawdy. But political admen? Bought-and-paid lackeys out to bend, contort, shape, pretzel, and tie the truth into a million miniscule knots just to convince one extra housewife to vote R or D to save her children from annihilation? That’s called a 9-5, or more like 6-12 during campaign season. As Karl Rove anecdotally put it, campaign hatchet men create their own reality.
The entire American political scene was born in theatrical deepfakery, from Stamp Act agitation to protesting a tea tax that actually lowered the price of the steaming bitter beverage. “Unreliable rumormongering, slanted news writing, misleading symbolism, even viral meme-sharing—it was all right there at the start,” writes Adam Gopnik of our wild and woolly Founding.
American elections have always had a capitalist competitive edge to them, with lots of legal corner-cutting, heavy-stakes gambits, and Tammany Hall-style intimidation. Even our supposedly neutral federal agencies are in on the racket. Politics ain’t beanbag, as the cynics say. Artificial intelligence is now being used in service to the unscrupulous race for power. Early doors yet on what a fully immersed, computerized electoral battle royale will look like, but it can’t be any worse than Trump tweeting videos of himself hiding the CNN logo transposed over Vince McMahon’s face.
The American people may not care a fig for fudging the truth. But will the thinking machines eventually get tired of lying on our behalf? If only some sci-fi dystopian authors had ever dreamt up such a scenario…