I don’t know. I asked chatgpt where I can get the most calories for my money when eating fast food, tempered by the qualification that the food is delicious. It very confidently and without any wishy washiness told me I should go to Wendy’s and get a burger. I highly doubt this is actually the best answer to my question, but I bet lots of people are asking similar questions. It made me actually think about going to Wendy’s. It read like an ad, in some ways, but some part of me took it in and tried to believe it because it felt much more similar to a recommendation from a friend (which I would for sure listen to) than an ad (which I almost always tune out—the louder and more obnoxious the stronger I look away).

astroturf abounds in text online. Sometimes its very obvious, sometimes not. I don't doubt it'd be effective, especially as people grow increasing dependent on AI and their ability to think and reason for themselves decline. There's concern that's already happening to people who use AI as it is and I doubt targeted, purposeful manipulation of the masses in exchange for cash will help that any.

> Subtle product placement and propaganda is not easy in text

Reddit.