I stumbled onto Synthesia today and made this quick video where I edited the templated script to use my name.
The process was simple and after submitting the request to create the video, a link to the page was sent to my email with the subject "Your AI video is ready".
For those who signed up, there are several actor/actress to choose from with capabilities to speak more than 50 languages.
She is talking with another AI on Youtube
Subsequently, I went to look for similar video on Youtube and found one where 2 AIs were talking with each other as though they understood contents and context.
This is both amazing and scary.
It is amazing because there are so many use cases for this capability, such as new broadcasters, characters in video games and movies, insurance advisors, corporate video presenter; just to rattle off a few at comes to mind immediately.
At the same time, it is also scary because the realism (and foreseeable increasing realism) could be used for crime and devious intents. Deepfakes impersonation is the tip of this ice berg.
Early stages of Andriod development
This seems like the early stages of andriod development. This AI capability will eventually work in conjunction with Hi Resolution 3D scanning, Robotics, and 3D printing of silicon masks and suits and the pieces of the technology puzzle while developing separately are coming together. Interestingly, about 3 weeks ago, Nvidia released a video where they show how a virtual Jenson was made for their GTC event.