MeTube
Moving forward five years, we’re going to hit a revolution in entertainment where what we’re watching is tailored specifically for us. It will be the funnest, most amazing content we’ve seen and we’ll think about how we lived before such an exciting era.
We have to start where we’re at now. Today, we have fairly elegant deep fakes that can fool a majority of people if they didn’t know better. We also have GAN-based image generators that can create almost any style of image within seconds. There are also others that can start with an image and edit it to fit a different scene.
We also have the very first animations coming from DALL-E. Combine this with indistinguishable speech synthesis and we start to build something that is compelling. You can have moving pictures, speech, and music, but you still need a plot. This is where GPT3 and its peers are going to have a big impact in changing the way these videos can made.
The first all-AI created movies are likely to come out within the next year. They’ll be OK. Better ones will start coming out the year after that. But it’s what happens next that’s exciting. With reduced costs in using the technology and better capabilities (and more realism), the videos will be able to scale infinitely. As we scroll or even while we watch a video, the video might change in real time to keep us entertained.
The result: there’ll be endless drama and it will be impossible to spoil a show.
