Raw LLM Responses

I want to start by steel-manning your position, because I think parts of your critique are valid and worth taking seriously. You’re right that current generative AI models were trained on large datasets that include copyrighted works and non-consensual artist labor. You’re also right that prompting does not develop traditional, transferable art skills like anatomy, composition, or draftsmanship. And you’re right that mislabeling AI outputs can harm working artists socially and economically, especially in an environment where accusations of “AI use” are now weaponized. Those are all real issues. But where your video falls apart is what you build on top of those points. First, you repeatedly slide from “this system has ethical problems” to “therefore anyone using it is illegitimate as a creator.” That’s a categorical leap you never justify. Historically, nearly every major creative technology displaced prior labor, embedded earlier skill, and reshaped authorship. Photography, digital painting, 3D rendering, collage, sampling - all relied on prior human work and all replaced jobs. What makes AI unsettling is its scale and speed, not that it’s morally unique. Second, your definition of “creative agency” is so narrow that it becomes circular. You define real art as "labor-intensive, technically embodied execution," then conclude AI users lack agency because they don’t do that execution. But agency isn’t binary. Choosing constraints, iterating, rejecting outputs, directing composition, and deciding what counts as finished are all real creative decisions, even if the execution is delegated. You can argue it’s less agency or a different kind - but claiming it’s none at all just isn’t defensible. Third, the claim that AI cannot produce anything “new” because all elements exist in training data is just factually incorrect. If taken seriously, it would invalidate human creativity as well. Humans also recombine learned techniques and cultural motifs. The meaningful distinction isn’t novelty versus non-novelty, it’s the intentionality, the embodiment, and the responsibility - none of which you actually analyze in depth. Fourth, the critique of Shadiversity drifts from methodological criticism into motive attribution. Instead of saying “your framing overstates your contribution” or “your terminology confuses authorship,” you imply avoidance of growth, disrespect toward artists, and self-inflation. That weakens your argument. You don’t need to psychoanalyze someone’s intentions to critique their claims, and doing so makes the video feel moralistic rather than rigorous. Finally, your energy is aimed almost entirely at individual users instead of the systems that created the problem: dataset governance, licensing, corporate incentives, and regulatory failure. Artists don’t need more gatekeeping between “real” and “fake” creators. They need structural solutions. Targeting end users while leaving the infrastructure untouched might emotionally satisfying, but in the end, it doesn’t actually solve anything. In short: while your ethical concerns may be considered legitimate, your conclusions grossly overreach. You collapse nuance into absolutes and define art so narrowly it excludes entire accepted creative disciplines (like collage), and personalize what is fundamentally a systemic issue. The result isn’t a strong defense of artists - it’s a rhetorically sharp but philosophically brittle take-down that convinces people who already agree and alienates everyone else. That’s why the video doesn’t hold up, even when its starting concerns are real. Sad to have wasted an hour and a half on this.