You’ve probably heard of a “straw man,” a fallacy in which you misrepresent an opponent’s argument so it’s easier to refute. Put up a false impression of what they’re saying, and then knock ’em down. We all do it.
But what we all probably don’t do is create a full-blown AI-generated deepfake of our opponent and then sincerely try to argue with it for nearly an hour. And we probably aren’t running for public office, either.
Credit where credit’s due, though. John Reid, a Republican candidate in the Virginia lieutenant governor race, went ahead and did just that, seemingly not stopping to ask questions like, “will this make me look insane?” or “does this count as spreading misinformation?” — because he is, you must understand, on the cutting edge of the AI straw man.
Reid streamed the 40 minute long stunt on his YouTube channel, and it’s every bit as absurd as you’d expect.
After boring us with nearly two minutes of text explaining that Reid’s Democratic opponent, state senator Ghazala F. Hashmi, repeatedly turned down his requests for a debate, thus forcing Reid to pursue the “best-possible alternative” of using AI to replicate Hashmi, an AI debate moderator pops up on the screen in the form of a still image, and introduces them both in a bland voice.
Both candidates give opening statements, with the AI Hashmi going first so Reid can get in a few easy dunks.
At long last, a wideshot reveals Reid dutifully standing at a podium and gazing at the podium across from him, where his opponent — a TV screen playing the AI Hashmi’s responses — sits equally resolute.
This is, without hyperbole, the best thing that's ever happened. https://t.co/hCaUy4wYvy pic.twitter.com/aZfz1dAFhH — David Weigel (@daveweigel) October 22, 2025
What does this say about the state of American politics? Our analysis: that things are very strange indeed.
“It’s not any worse than having an AI Trump piloting a plane and dumping on the public,” Virginia political strategist Bob Holsworth told the Washington Post.
... continue reading