Why don't we read AI-written text with the same skeptical eye we use for ads?

This explores whether we can learn to take AI-generated content with a grain of salt, the way we already shrug off advertising, and what stands in the way. The short answer from the collection is that we haven't done it yet, and the reasons are worth knowing. We discount advertising because we know who is speaking and what they want. Every established source of public speech, whether ads, press releases or campaign speeches, comes with a built-in reading posture that filters how we receive it. AI text arrived too recently and changes too fast for any such posture to form, so it circulates without the protective skepticism we apply to 'interested' speech by reflex How do we learn to read AI-generated text critically?.

The advertising comparison also breaks down at a more basic level: you can't discount what you can't spot. Ads are labeled, formatted and placed in recognizable ways. AI content isn't. A review of 30 studies found that people tell AI-made text, images and voices apart from human-made ones at roughly coin-flip accuracy, and that accuracy hasn't improved as the models have gotten better Can people reliably spot content made by AI?. Even when people know they're looking at AI output, they often stop checking it. Fluent, confident answers build false trust, and verifying them takes effort, so users accept AI claims without challenge in large majorities. The collection calls this 'cognitive surrender' When do users stop checking whether AI output is actually backed?.

There's a less obvious problem too. The discount on advertising works partly because ads look like ads: we can recognize the sameness of the genre. AI may mass-produce similar content that feels personalized, so the uniformity is invisible to each individual reader. That makes it a quieter version of the old 'culture industry' critique of mass media Does AI homogenize culture the way mass media did?. On social media, AI posts collect likes through thorough, confident phrasing but don't invite replies. They gain the look of social proof without the back-and-forth that used to earn it Why do AI posts get likes without inviting conversation?. That pushes out the human voices whose reputations the platform was built to grow Does AI content displace human influencers on social media?. The damage is to the conversational texture of the medium, and it sits below where moderation or fact-checking can reach Does AI threaten social media's conversational function?.

A historical framing helps explain why this is hard. Print culture fixed knowledge as a stock of authored, attributable texts. AI returns us to something more like oral 'flow', but without the speaker or giver who once anchored trust in what was passed along Is AI returning knowledge to flow-based economies?. Our discount on ads is aimed at a speaker's motives. AI content often has no clear speaker to be suspicious of. And it arrives faster than human judgment can evaluate it, while the tools for checking it are increasingly AI-made too Can AI generate knowledge faster than humans can evaluate it?.

The unexpected takeaway is that the ad analogy may point to the wrong fix. A cultural discount on advertising targets *who is speaking and why*. For AI, the more useful skepticism may target *missing conversation*: content that is polished and complete but invites no reply and has no one behind it. The collection doesn't yet describe a working cultural practice for this. It does make clear that labeling and detection alone won't build one.