I had an epiphany about synthetic media back in December 2017. In that epiphany, I came to the conclusion that we were, at most, ten to fifteen years away from the day that artificial intelligence could reliably synthesize any sort of media, and that we were likely only five years away from the first big public demonstrations of this technology. Five years after that epiphany, we lived in a world of DALL-E 2, Midjourney, Stable Diffusion, ChatGPT, HarmonAI, and much more. As I anticipated, these were still fleeting examples of what I imagined, with the machines incapable of maintaining long-form coherency. However, with talent, one could utilize these tools effectively. However, my epiphany was fairly limited and conservative for the near future. I predicted that GANs— generative adversarial networks— would drive this revolution. I presumed that movies and 3D procedural generation would not truly begin taking off until 2027. And I even said that none of these tools would be available on your desktop within five years— and perhaps I was right in some regard since, besides Stable Diffusion, none of these are locally run. Perhaps I simply underestimated the Cloud to that end. It might seem in retrospect that I was being vastly too conservative. In truth, I actually thought myself far too liberal for a time, especially in 2018 and 2019 when so little visible public progress in synthetic media occurred. When I created a follow-up to that epiphany in 2022, I felt the need to refine some of my thoughts and be a bit more deliberately liberal. Now, even those predictions I outlined not even a full year ago are starting to come off as increasingly conservative as more capabilities are derived from diffusion models. Capabilities I presumed to be reasonable for 2027 now seem likely by next year. But my central point back then wasn't to focus on the technical details, but rather the emerging capabilities, the "democratization of creation." Even back in 2017, my understa