Throughout the creation of the content that formed the basis of the Elyssaverse, one of the key parts has been the use of the AI video generation platform Sora to “bring the characters to life” in a consistent manner. OpenAI’s video generation platform allowed for the characters to be saved and “called up” at will with defined appearances, personalities, and manners of speaking. In addition, the company had negotiated a deal with the House of Mouse itself, penning a multibillion-dollar deal with Disney to allow their roster of characters to be used in Sora videos. Imagine my surprise when I discovered this announcement from the company when browsing Reddit last night.

This development represents a major shift for this website project. I had always intended for this site to serve as both an archive of my existing work and a “springboard” for the creation of new content for the story. That archival element has now taken on an entirely new meaning. I had not originally intended to preserve everything on this site, but now, my opinion on that has shifted in the wake of this announcement. I am currently in the process of downloading all of my creations from the Sora app, and I intend to archive every single one of them here, good and “bad” alike. At least here, there will be a complete collection of one creator’s entire experience with the at-times quirky platform.

https://www.bbc.com/news/articles/c3w3e467ewqo

As this BBC article states, the Sora platform was a huge money pit for OpenAI as well as a drain on their computational resources. Shutting down Sora will allow them to focus more heavily on the lucrative enterprise coding market that is currently dominated by their competitor Anthropic’s Claude platform, as well as apparently re-directing Sora’s resources into various robotics projects, presumably in preparation for the company’s rumored IPO.

So, what does this development mean for the future of Elyssaverse video content? I intend to explore some of the other available video generation services once I break out of this bout of “writer’s block” that has spurred the creation of this site in the first place. I intend to test Midjourney, Stable Diffusion, and Gemini’s video creation features to see which can best capture the “feeling” of the original Sora content (I would also look into Seedance 2.0 if that platform ever fully comes to the US). This development also means that I will have to alter my existing templates to remove elements like character tags to make them more universal, as well as creating new templates that will hopefully create some manner of consistency in the characters’ appearances. In other words, making sure that “Gemini Elyssa” or “Stable Diffusion Elyssa” still looks like Elyssa at the end of the day.