Advances in AI Video

The latest in AI music video developments

Now that AI video is starting to catch up to the excellent state of AI music, the obvious question is what can be done in terms of combining the two. This first video is my first successful attempt to turn a Soulsigma song into something that is halfway-watchable; it’s far short of a professional standard, of course, but it actually manages to tell a coherent story, at least as coherency is defined in the world of the music video.

This is a new rock mix of IF YOU HAD A TIME MACHINE. What’s featured here is the updated version of the video; you can see the original version of it here.

Even keeping in mind that I had absolutely no idea what I was doing, to the extent that I literally didn’t know how to combine two clips together and had to ask for an explanation, the most remarkable thing about the original video is that it was constructed from a single low-resolution photograph taken in 1993.

However, you can already see the technical improvements in both the updated version and the following work in progress. Now, I’d previously promised a reader on another site to explain how I made that video, but I since I’m rapidly learning how to do things better, I’m going to save that for a future post here. In the meantime, though, you can see some of those improvements in this work in progress, which is a video for a Soulsigma song called Chiba City Blues from the next album.

If the improvements continue, I’d estimate that we’ll have functional video quality that approaches the current level of functional audio quality within 24 months. It will probably be possible for video experts to get really good results sooner than that, but the real challenge is likely to be more in the area of user interface than in video generation.

And, of course, remember that music videos are just the baby steps toward full AI movies and television. That will take a while longer, but it will be sooner than anyone in Hollywood is ready for it.