Human beings are prone to mistaking local consensus for global consensus.
Most people never notice this process happening because they're never pushed far enough outside consensus to see its edges. Increasingly, we encounter flattened symbolic versions of one another first, mediated through our local cluster's narratives.
AI Bro. Luddite. Victim. Abuser. Grifter. Shill. Fascist. Communist. Capitalist.
Once someone is compressed to a symbol inside your local consensus reality, their actual interiority stops mattering. Their motives become obvious, their beliefs predictable. Their humanity becomes secondary to their usefulness as a representative of the category you've assigned them to.
This was prompted by a viral thread on Bluesky from an animator asking essentially: if AI were creatively revolutionary, surely we would have seen some huge grassroots breakout phenomenon by now. Something analogous to Flash animation in the 2000s. Something undeniably born from the technology itself. It was, as best as I can tell, a response to an animator being announced as part of a project at Amazon to experiment with AI use in processes, right before he left facing a wave of backlash from people who viewed participation in AI projects as morally unacceptable.
How would you even identify something made in part or whole with AI if creators had strong incentives to hide the workflow? They see what happens to people who disclose it.
That's not proof that such successes exist. It doesn't answer the challenge entirely. The Flash era was visible precisely because people openly disclosed their tools. If today's equivalent is partially hidden, that very invisibility makes it harder to establish as a cultural phenomenon.
But people are making things with it.
Experimental animation. Hybrid creative workflows. Tiny bespoke utilities. Collaborative riff spaces where people rapidly iterate on each other's ideas. Weird little micro-scenes full of artists and tinkerers.
Some are open about it. Many are not.
And I don't think I'm fully outside this dynamic either. I can feel the pressure field around disclosure. The constant ambient demand to frame AI as contamination. To pre-edit yourself around anticipated reactions. The demand to disclose AI use and paint a target on your back.
Algorithmic timelines have made profound social isolation available at scale. You no longer need the means to physically isolate yourself. Your information environment can do much of the work for you.
This is how you get tech CEOs who genuinely believe everyone loves AI and highly toxic online communities full of people who genuinely believe everyone hates it.
When students boo a CEO evangelizing AI in a speech at their school, some observers see proof of widespread anti-AI sentiment. Others see students rejecting the person delivering the message rather than the technology itself. Given how widely these tools are already being used, neither interpretation seems obviously sufficient on its own.
The point isn't which side is correct. The point is how readily people assume they already know. It's how people convince themselves harassment is activism. Or that ruining a stranger's livelihood is morally justified because the stranger belongs to the approved out-group.
We are a long way past that video of Will Smith eating spaghetti in only a few years and outputs you can't identify as generated will only get more accessible to less sophisticated operators. You probably don't know all the shows made in Flash, and you won't know everything made with AI until it's just another normal technology and we find out in retrospect.
The anti-AI backlash and the ongoing atomization of culture are absolutely headwinds against any openly AI-heavy production achieving broad cultural legitimacy right now. Maybe you see that as a good thing, but the backlash also distorts our ability to measure what is actually happening. We don't know what we don't know.
People stop saying what they think.
They stop disclosing what they use.
They stop experimenting publicly.
They retreat into smaller and smaller enclaves.
And everyone thinks they are describing this phenomenon from safely outside of it.
They aren't.
Even people I know who are deep in using AI for productive work struggle to believe it's useful for anything other than coding until I show them examples. Social media bubbles are real, and they're more atomized than we tend to think. It's not left vs right, it's micro-cluster vs micro-cluster.