lol. AI can only re-assemble what it has been fed (and often, badly), so AI is the very definition of a reproduction of dominant ideas (with hiccups).

The idea of this dubious paper by my esteemed colleagues relies on the insane idea that AI knows "truth", and on the even more insane idea that people will beliebve somethinig -because- it comes from an AI, thereby regurgitating all the sci-fi ideas of infallible machines, "because math", like the old insight than anything seems more credible if you put a number and a graph next to it.

"Amid growing threats to democracy, Costello et al. investigated whether dialogs with a generative artificial intelligence (AI) interface could convince people to abandon their conspiratorial beliefs." In other words, amid growing threats to the dominant narrative as put out by the media, the researches try to see if a bot is seen as more "neutral" than a TV presenter and can be utilized to harness the credibility of "math" and "computers" and "graphs" by being math-generated text.

The problem is that AI has destroyed that image. When Google gives you an AI answer, you go look for a human one, as we've learned very quickly that the AI can't be trusted. This will backfire.