There are technologies that can automatically adjust videos for eye contact today, so I imagine something similar could be implemented for this later on.