Web Summit: future selfie will be an intelligent avatar

The “ultimate selfie” would be an avatar that would live on after you, with its own “memory” and ability to interact with your children and grandchildren, an augmented reality scientist told the summit.

Jacki Ford Morie, founder and chief scientist at The Augmented Traveler, said that while the technology did not exist now to create such an authentic "living" avatar, it might do so in 10 years or a little longer if funding was put into it now.

In a talk on the Society stage, Ms Morie said the sensors people were wearing now were getting smaller and smaller and would be woven into our clothes.

“Every aspect of our physiology, of our biology can be captured by these in the near future. I think that’s one trend we have to look for.”

The ways in which we were able to capture our physical appearance were also becoming much more sophisticated and could even capture layers of skin, she said.

One particular device the film industry used could create a hyper-realistic double of an actor, for example.

This was something that was coming down in cost and in access and in the coming years people would be able to capture themselves and their children in full 3D and to be able to look at the captures and really know what the person was like.

In the virtual world, people were increasingly using avatars to create digital forms of themselves and social communities, Ms Morie added.

“The lines between what’s physical and what’s digital are blurring. What’s it going to be like in 100 years?”

She said she believed people would be able to live in these digital realities and to move between the physical and the digital in ways they could not even imagine today.

Already there was an artificial intelligence company called AI that could scan people in 3D while they were moving. Demonstrating the technology on screen, Ms Morie said you could put the scans in a virtual space and walk around the person while they were talking.

“I would like to have an avatar that not only is a recording of myself but actually embodies more of my intelligence. This is what I’m calling the ultimate selfie. But to do this we need new forms of artificial intelligence.”

She said she really wanted digital entities to learn. She wanted new forms of AI which were not only smart about what they knew, but were smart about the human that’s interacting with them.

A virtual human called Sim Sensei had been created to interact with humans and scan their faces and body movements, the audience heard.

“This virtual human knows exactly what your emotional state is, because it is scanning the facial expressions, the eye-tracking and the body movements.” It could also tell whether someone was agitated and how the real human was reacting to it.

“They adjust their body movements and the things that they say to build up a rapport with that physical human who is interacting with them, and this is really cool. But they don’t remember you from one form to the next.”

Very sophisticated neuro-computational models created by a scientist in New Zealand, which integrated theories of social and emotional behavioural systems. He used them to create what he called artificial nerve signals to drive the facial muscles in a virtual reality "baby". It could be happy or sad and would enter into a sleep state if you didn't interact with it.

Ms Morie said she really wanted her own digital avatar or digital double to learn how she behaved and to understand the experiences she had as that avatar.

“It would live on after me so that my great-grandchildren could come in and talk to me after I’m gone, because this digital entity will really behave as I behaved because it learnt from me throughout my life.”

Asked about the potential for illicit uses of such artificial intelligence, Ms Morie said people could potentially masquerade as someone else or steal their digital double and “put some bad memories in there”.

On whether it might create a problem for younger generations in dealing with death, she said she believed the idea was comforting and that leaving behind a photo or video was more of a “dead, passive thing”.