Twenty-sixteen is going to be a pivotal year for virtual reality. Already, the New York Times used VR to cover the emotional aftermath of the terrorist attacks in Paris. Apple has implied an interest in VR in its most recent earnings call and made key hires like that of Doug Bowman. And most critically, consumer launches of products like the Oculus Rift will finally happen.
The rich potential of VR is best described by one of its leading auteurs, Chris Milk. He has documented a Syrian refugee camp for the United Nations. He has covered the Millions March against police brutality for Vice. And much more. In his 2015 Ted Talk, Milk said of VR:
It connects humans to other humans in a profound way that I've never seen before in any other form of media. And it can change people's perception of each other. And that's how I think virtual reality has the potential to actually change the world.
Enabling this "empathy machine" as Milk refers to it is a powerful aspiration. And one I think many creators could believe in. Unfortunately, as with many emergent technologies, VR has a bias baked in: It's sexist.
An age-old example of gender bias in technology is the humble air conditioner. When standards were put forth in the 1960s comfort levels were set based on tests conducted on 40-year-old men. That's problematic given men and women have different metabolic rates. This is why women are more likely to be cold in offices. Men are freezing women out of the workplace.
Another more recent example proves the problem has followed us from atoms to bytes. Research by Carnegie Mellon in 2015 outed Google algorithms as "accidentally sexist" too. The CMU study revealed that Google was displaying more prestigious job ads to men. Algorithms created largely by men are exacerbating the gender pay gap issue. Even as some governments are forcing companies to finally publish data on the topic.
Based on that pattern it should come as no surprise that VR suffers from much the same. Motion sickness in VR has plagued the format since its inception. Women have shown a greater tendency toward VR-induced nausea than men. But why? It's all about unconscious bias and technology’s notorious self-selection bias.
In brief, there are two main methods of suggesting distance in most modern VR systems. Motion parallax and shape-from-shading. At its simplest, motion parallax is a play on perspective. Shape-from-shading is about how light sources interact with objects and your point of view. They’re conflicting depth cues.
Most systems favor motion parallax as it's easier to program and render. As a result, it's the default model used by most. And guess what? A study by danah boyd of Microsoft Research shows that men prefer motion parallax while women prefer shape-from-shading. Men are literally sending the wrong signals to women’s brains!
Another possible explanation was presented at the National Academy of Sciences/National Research Council Committee on Human Factors in a paper by R.S. Kennedy and L.H. Frank in 1983. The pair’s research suggested a different biological factor at play. They believed the higher incidence of simulator sickness in women was due to a wider field of view. Other researchers pursued the topic in 2006 and noted the severity was even higher in older women.
Despite all that direct and related research, boyd became a flashpoint for gamergate-style blowback. On Oculus's own developer forums even. You can read it here at your own peril. Peter Andrew Nolan, a noted Men's Rights activist was the ringleader of that hot mess. A sample:
- boyd's opinion was tantamount to classic "MSFT sabotage tactics"
- "her research was sloppy"
- "women are not trustworthy"
- "It's not VR's fault! It's biology!"
- "feminists and their mangina lackeys are going to attack anything that liberates men from the clutches of women."
Defaulting to masculine factors isn't just a software problem. The bias exists in the hardware as well. Adi Robertson, who covers the VR beat at The Verge, recently covered a host of "atoms" problems as well:
- headsets that barely tighten enough to fit
- lenses too far apart to focus properly
- oversized motion capture rings
- ill-fitting haptic suits
As Robertson points out, when you're in the early stages of development, decisions like these are perhaps understandable. But with products close to market it becomes disheartening that varied sizes aren't available. It's astonishing especially at an event like CES. Those early decisions build a weak foundation. As a result, concerns like this end up addressed in third or fourth product generations, if at all.
The key lesson to take away from all of this is simple. The only way to beat the sort of self-selection bias at the root of this problem is to get more women in VR. To join luminaries like Carolina Cruz-Niera, Brenda Laurel, and Margaret Minsky. Thankfully there are groups like Virtual Reality Girls driving this mission forward. And events like VRUK Fest that create friendly environments to probe the issues further.
Silicon Valley brims with hiring initiatives to tackle gender bias organizationally but progress is slow. In the meantime, the onus rests on existing developers to practice inclusive design. Diversity is a design problem.
How can we create Milk's "empathy machine" if we ourselves aren't more empathic? How much of VR's potential has been held back by focusing on the male default? What might we create that we wouldn't have before?