Artists Have Already Made Up Their Minds About AI

The most telling sign that a cultural debate has matured isn't when people start arguing — it's when they stop. The conversation around AI and creative work has reached something close to that point, at least on Bluesky, where sentiment against AI-generated art sits at nearly two-thirds negative, nearly triple what it was just days ago. But this isn't a panic spiral. The posts don't read like people encountering something new and frightening. They read like people who have already decided. "AI art has gotten steadily worse since Secret Horses," one user wrote, pushing back against the familiar techno-determinist argument that resistance is futile. The post drew more engagement than almost anything else in the thread — not because the claim is empirically verifiable, but because it articulates something many in creative communities clearly feel: that the promise of generative AI as a creative force has failed on its own aesthetic terms, independent of every other grievance attached to it.

Those other grievances are still very much present. Copyright runs through this conversation like a fault line — the Korean Cartoonist Association held a webtoon forum specifically to address it, a U.S. Copyright Office filing in the Jason Allen case is circulating among legal observers on Bluesky, and the ambient anger at training data scraping ("I'm surprised they're bothering to pay for anything") has calcified into a background assumption rather than an active outrage. What's notable is how legal anxiety and aesthetic disdain have merged into a single posture. The people denouncing AI stock art flooding markets and AI-generated game assets are making the same moral argument as those tracking courtroom filings — that something was taken, and that the outputs aren't worth what they cost. Meanwhile, the two plainly pro-AI voices in the sample don't attempt to rebut any of this. One sells "designer-led AI visual assets." The other argues that copyright is "an innovation blocker" and that "data must flow." Neither engages with the creative community's actual complaints. The camps are not in dialogue.

The sharpest divergence isn't between artists and technologists on Bluesky — it's between Bluesky and arXiv. News coverage of AI and creative industries is running deeply negative, and Bluesky mostly tracks it. But the small cluster of arXiv papers touching this beat scores positive, reflecting a research community that is genuinely excited about generative tools for creative workflows and tends to frame copyright and labor questions as tractable engineering or policy problems. These two worlds — the researcher testing a new LoRA fine-tuning approach and the illustrator watching their stock art market collapse — are having entirely separate conversations that happen to share a vocabulary. "Generative AI" means something different to each of them, and that gap is widening rather than closing. What the Bluesky discourse reveals, more than anything else, is that creative communities are no longer waiting for that gap to close. They've decided where they stand, and they're building their identity around it.