Evan Ratliff’s Shell Game Podcast Delves Into AI’s Threat to Our Humanity
Evan Ratliff keeps trying to disappear. Fifteen years ago, he evaporated from the grid in a grandly designed social experiment for Wired, leaving behind only a gutsy, playful, disrespectful challenge: Find me. Readers organized in the nascent social web. Ratliff dodged and redirected with new identities and digital distractions. In a dramatic dash, he was nabbed by his bounty hunters, ending his run as a ghost in a piece of precedent-setting performance art.
All of which is to say: Mr. Ratliff has had a bit of practice for his new gig. But this time, in his compelling and unsettling podcast Shell Game, the ghost is in the machine. Still, it remains his. Using emerging AI technologies with existing telecom capabilities and the self-effacement of his Wired disappearing act, Ratliff sculpts a convincing digital twin — a voice agent that sounds like Evan, responds like Evan, and maybe even thinks like Evan. Whether that last bit is true is, in many respects, the essence of the experiment.
Ratliff’s AI laboratory is big, and he quickly makes it his playground. With only six episodes to create and resolve an arc, Ratliff starts at a sprint: “Fake Evan” is talking to customer service representatives within minutes. By the season’s midway point, Ratliff’s digital twin has bested scammers. And when we approach the final act, therapists are fooled, loved ones are unsettled, and listeners are agape. Our astonishment is derived from the AI capabilities and the underlying critique of a system that hurtles toward a precipice of exploitation and absurdity–unregulated and unabated.
Something about Ratliff’s work functions like a George Carlin set: you’re laughing at the joke so hard, it’s only on the drive home that you realize the world’s been undone. There are plenty of good setups and dismounts: leading a scammer into a gratifyingly time-destroying cul de sac and listening to Fake Evan work out a misunderstanding with another bot. Ratliff’s restrained production –privileging an intimacy and minimalism instead of the “Radio Lab” overwrought –provide ample time to the experiments, adding to the eerie, voyeuristic quality of each episode. Ratliff’s unassuming to his core, a podcast personality and hesitant self-promoter who somehow feels uncomfortable at the notion of his own publicness while devotedly creating more of it. The tension works, making his grappling with what the project creates and what it means all the more persuasive. Alarmists are shrill. Our techno-optimism froths despite the whistleblowers. Perhaps it is formats like Shell Game, where gifted journalists can awaken our social consciousness.
One moment, in particular, is hard to shake: Fake Evan calls a close friend of Ratliff’s, who believes not only that it’s Ratliff himself but that Ratliff must have suffered a kind of mental break. The friend himself is broken at the thought. Frankly, the listener never fully recovers, either. Is this, one wonders, the catastrophe that awaits us? The fragmentation of our identities, distributed over so many horcruxes, the dehumanization that attends the disembodied, unattended self?
In a work that echoes across the AI canyons of our moment, Jean Baudrillard wrote in Simulacra and Simulation of a coming time when the distinction between real and artificial becomes irrelevant. A friend can’t tell your voice agent from yourself. A voice agent performs journalistic interviews. You receive your last call from a real human sales agent. A boy in Florida becomes obsessed with a chatbot and kills himself at her prompting. Soon, all of us are caught in a perpetual artificiality, what Baudrillard described as “an uninterrupted circuit without reference or circumference” or, more achingly, “a machine for making emptiness.”
