In keeping with this year’s theme of place, purpose, possibilities, we explore the possibility of using AI and digital people as both educators and educational surrogates. Imagine if your lectures could be updated online by an AI engine.  Perhaps altering a recording of your previous lecture, or updating the video to show you teaching new material? What benefits could this technology bring to students?  Would a girl learn maths better online from a digital avatar that looks like her same age?  Or what if that digital girl was a digital version of herself?

Imagine, if you could leave the University, but still teach in a virtual classroom? If the University owns the copyright of your material, will you soon have to sign away your rights to your likeness to allow for future use?

We are at a nexus point, a perfect storm of new techniques, faster GPUs and machine learning. As this technology moves out of the research lab and beyond the entertainment industry, will a professional version of Deep Fakes change the education landscape, especially for wide deployment of individualised programs?

At the Motus Lab at The University of Sydney we have been building digital humans and researching their use. Instead of these trend worrying undergraduate students, our research shows that they are not alarmed at all by the advent of AI-driven agents. While some futurists wax lyrical about the ‘Robot Uprising’ or mythical ‘singularities’, in this presentation we examine the current best practice applications of this technology in a very grounded and concrete way. We move past the hype to outline a framework of practical implications and show real-world, state of the art implementations in this rapidly advancing area of research.

Find out more about Mike Seymour

Recommended Articles