“Everything you can imagine is real.” ― Pablo Picasso
As Picasso’s quote above serves to illustrate, the capacity of the human imagination to generate alternate realities is essentially without limit. Yet, for visual artists and programmers working in social VR, there is a wide range of practical design choices that must be made in order to create the most suitable conditions for human communication and interaction. Can avatars look too realistic? If individual movements can’t be tracked, should they be rendered anyways based on other available inputs? Should eyes be rendered with or without pupils?
In a recent “Fireside Chat” with our CEO Eric Romo, held live within AltspaceVR to a global audience, Romo discussed the scholarly research and design principles behind avatars in social VR, uncovering some key learnings for anyone wondering how and why these virtual experiences actually look and feel the way they do.
Ultimately, Romo revealed that what we consciously believe to be the ideal design in virtual reality is different than how we act once we start engaging with one another in real time, leading to some unexpected results for current users and for future iterations.
Filling in the “Communication Chasm” of Social VR
Romo began his talk with a slide covering the spectrum of different communication options that fall beyond real life “face to face” interactions. From written letters to text messages, from phone calls to video calls, the depth of communication, level of interactivity and emotional connection varies due to the different physical signals available for each medium.
So, what happens when we start to communicate “face to face” with avatars in virtual reality? What are the various verbal and visual signals we should expect, and how do they influence the experience? Suddenly, the level of design variation within that virtual environment creates an entire “chasm” of communication implications that programmers need to fill.
And, once you begin walking around AltspaceVR, our users slowly become aware of all of the non-verbal cues that come into play, all in ways we’ve never before experienced. Consider the following, and try to imagine how avatar design would be impacted by each:
- Eye contact
- Personal space
- Body language
- Direction and volume of voice
- Facial expressions
All of these factors raise important questions about how we will all behave and interact together in virtual reality, and each must be accounted for when designing avatars.
When Real is Too Real: The “Uncanny Valley” Problem
Looking specifically at avatar design, one of the biggest issues Romo discussed is something known as the “Uncanny Valley.” Originally coined by the robotics professor Masahiro Mori in 1970, the hypothesis is that as avatars begin to look and act more like humans, the improvement eventually reaches a tipping point where we feel an innate revulsion to the avatar. Why? We can tell something just isn’t quite right, creating a negative emotional response to the creepy “being” in our midst.
In order to avoid the uncanny valley, you either have to continue making the avatar more and more human, or you have to pull back and intentionally make the avatars less human in one or more aspects.
Thus, while we might believe that interacting with a “more realistic” human avatar is ideal, research shows that once we are actually in the presence of one of these “not quite human beings,” our physical emotional response tells a very different story. As technology continues to improve, it’s reasonable to envision a future where we end up making it through the uncanny valley. Though, that undoubtedly will come with it’s own unexpected consequences. (If you aren’t thinking about The Matrix right now, you haven’t been paying attention.)
The Journey So Far, and Future Iterations
After discussing the foundational research behind avatar design, Romo walked the audience through our own avatar history to date in AltspaceVR. Since 2014, it’s been a constant learning process. As new iterations were released, our user feedback helped to shape the avatars in ways that minimized the creepiness levels and maximized social interaction and communication levels.
A highlight of Romo’s talk came at the very end when he began taking questions from the fellow avatars around the room. The result was a real time case study for everything he had just covered. Watching Romo’s avatar turn and face one of the participants and engage in a conversation feels surprisingly similar to being in a real lecture hall. You can even sense the energy of the other participants as they all gaze around the room wondering who would ask the next question.
By the end, it became abundantly clear that the users themselves are the most important factor in avatar design. And, if the energy at Romo’s talk were any indication, our users are leading us into a very bright future.
For more information on AltspaceVR’s check out their site here.