Unstable Diffusion will find another way to fund itself. From the article, looks like they may have already found a viable option. Artists will decry AI art fo

Unstable Diffusion will find another way to fund itself. From the article, looks like they may have already found a viable option. Artists will decry AI art for a few more months / years, then dust will settle, and we will still have ever-improving diffusion (or whatever the next iteration is) models. Slowly, people will realize that artists learn and copy from other artists much in the same way that current ML models do. Education about how these models work will improve, and the current mythology of "it's blatantly copy/pasting!" will subside, for the most part. We'll start to see a counter-protest against anti-AI artists who use tools like Photoshop or other digital tools, most of which have involved some amount of machine learning (e.g. the "heal brush" or "context aware fill") for a long time. AI art will start passing the "art turing test" at higher rates, and people will realize that whatever intrinsic spark / soul / whatever there is to art is mainly attributed to a narrative or story about the artist that exists outside of the canvas - and that the actual artistic contents of what lives on the canvas itself doesn't depend on a living, breathing thing if you make up a compelling narrative to support it. Artists displaced by this new ease of generating art will find new economic niches, and life will continue on.