Unpopular Opinion: The 'That's So AI' Insult Gives AI Too Much Credit

There is a new insult sweeping through group chats, TikTok comments, and school hallways, and it stings precisely because of what it accuses you of not being. "That's so AI" has become Gen Z's favorite way to call something fake, according to a recent report from The Up and Up newsletter. The phrase covers a lot of ground: a photo that looks over-edited, a friend acting robotic, a text that reads like it was copy-pasted. What it really means, the newsletter's Rachel Janfaza writes, is that something looks fake, that someone is being fake, or that someone is acting robotic, showing little emotion. Read the report here.

Over the summer, Love Island fans put the trend in lights when they nicknamed contestant Corbin "CorbinGPT" for his machine-like mannerisms. TikTok creators like Madeline Salazar have built followings playing a real-or-fake guessing game with their audiences, yelling verdicts at images of farm animals, mountain ranges, and bodies of water like it is a game show. The outrage is not always aimed at strangers, either: Zelda Williams recently pleaded with creators to stop generating AI videos of her late father Robin Williams. As one Yahoo report on the trend puts it, "clanker," a derogatory term aimed at AI and robots, has joined the vocabulary too. Kids raised alongside artificial intelligence were supposed to be its easiest marks. They are turning out to be some of its harshest critics.

That skepticism shows up in the data. For the first time since Pew Research Center started tracking the question, a majority of adults under 30 say the increased use of AI in daily life makes them feel more concerned than excited, according to the newsletter's summary of Pew's latest findings. The center's Wave 195 questionnaire, fielded June 22 to 28, 2026, put the question directly to respondents: does the growing presence of artificial intelligence in everyday life make you feel more excited than concerned, more concerned than excited, or both equally? The under-30 answer has flipped. See the Pew questionnaire here. Meanwhile, Gallup's 2025 "Voices of Gen Z" research found that 29 percent of daily AI users say the technology makes them feel anxious, and more than half of Gen Z adults who have never used AI say the same. Even the heaviest users are uneasy.

What the insult really means

Strip it down and "so AI" is shorthand for anything that feels mass-produced, thoughtless, or emotionally flat. And plenty of fully human culture qualifies. The corporate apology video read from a teleprompter is "so AI." The influencer who posts the same paid partnership as eleven other influencers in the same week is "so AI." The politician's statement that says nothing in four paragraphs is "so AI." None of it needed a language model. It just needed a person who did not care enough to be real.

There is a telling contradiction in how the insult gets used, too. Pew's December 2025 research on teens found that roughly 64 percent of U.S. teens say they use AI chatbots, including about three in ten who do so daily. These are the same teenagers yelling "fake" at AI-detector TikToks. Nobody says they cannot use the tools and criticize the output at the same time; that is normal. But the line between "AI I use" and "AI I mock" is drawn entirely around vibes, not substance. The chatbot that helps with your history homework is a helpful assistant. The image in your feed that you did not like is "so AI." The difference is not the technology. It is whether you liked the result.

The unpopular part: blame the humans, not the machine

Here is the part people will not like: calling everything "so AI" gives AI far too much credit. The slop people keep blaming on the machines is, in almost every case, made by people. A human set up the content farm. A human wrote the prompt, posted the fake product listing, and flooded the feed with synthetic junk to farm engagement. When a teen calls a blurry dropshipping ad "so AI," the accurate insult would be "so lazy." The machine did not decide to trick anyone. Someone with a spreadsheet did.

Researcher danah boyd, quoted in the Yahoo report, explains that humans instinctively look for the person behind a creation, and that kids encountering content with no visible author have developed what amounts to a "BS detector" for anything that feels inauthentic. That is a genuinely impressive piece of cultural self-defense. But a BS detector is only useful if it points at the right target. Right now it points at the tool, while the people mass-producing low-quality junk for clicks walk away clean. The insult treats AI like the scammer instead of the crowbar.

This matters because slop is not an AI problem. It is an incentives problem. Platforms reward volume over quality, advertisers pay for impressions rather than attention, and audiences keep scrolling. Put a person in that system and you get slop with or without a chatbot. The "that's AI" label lets everyone involved pretend the ugliness appeared on its own, like a stain with no one to blame.

None of this means AI is harmless or that the kids are wrong to be suspicious. AI-generated falsehoods carry real consequences, as one recent deep dive into a hallucination that rattled US-China relations showed. The Pew numbers showing under-30 concern outpacing excitement for the first time are worth sitting with. Young people live inside this stuff every day, and their discomfort is data. But if the goal is a culture with less fakery, the insult needs an upgrade. The next time something in the feed feels hollow, "that's so AI" should be aimed past the machine. Ask who made it, who paid for it, and what they wanted you to do. That is where the fakeness lives. The AI was just the cheapest paintbrush in the room.