Slider to control expressiveness

I read the Forbes article so I now know you’re definitely on the development team. How long will it last that you can actively listen to us fans of your service? What you guys have made is fantastic. There’ll be others that do the same thing of course, but brand loyalty and engagement goes a long way I think.

Assuming the model you’re running has a system prompt at the back end of our audio and image inputs, then it would be nifty if part of that prompt was tied to a slider. Subtlety will help with the realism I think. We humes know exactly what a realistic person looks like. Being able to dial in the level of animation will let us fine tune that realism rather than you guys trying to bake it into the model.

If I had a parameter list I’d add blink rate. A nervous person blinks more often than a calm person. And angry person may not blink at all. The blinking is my favourite part of the animation, but if it is also baked into the model then some of the realism for us as directors of actors is removed.

I’d also add head movement as a variable.

And eye gaze. A nervous person may not look back at the camera. They’ll look down and to the side. A confident person will be the way the model currently runs, being direct eye contact.

I haven’t read through the entire wish list for development yet, so I’m probably repeating what others have said. Plus you’re busy so don’t have to time for long comments. I’m just enthused about your service.

Animating my books.
Creating educational content.

Creating advertising/marketing avatars.

Video answering machine.

That’s four of a list that keeps growing in my head of what image + audio = animation can do.