Computer Scientist and Musician; Columnist, Discover Magazine
Here's a happy example of me being wrong. Other researchers interested in Virtual Reality had been proposing as early as twenty years ago that VR would someday be useful for the treatment of psychological disorders such as post-traumatic stress disorder.
I did not agree. In fact, I had strong arguments as to why this ought not to work. There was evidence that the brain created distinct "homuncular personas" for virtual world experiences, and reasons to believe that these personas were tied to increasingly distinct bundles of emotional patterns. Therefore, emotional patterns attached to real world situations would, I
surmised, remain attached to those situations. The earliest research on PTSD treatment in VR seemed awfully shaky to me, and I was not very encouraging to younger researchers who were interested in it.
The idea of using VR for PTSD treatment seemed less likely to work than various other therapeutic applications of VR, which were more centered around somatic processes. For instance, VR can be used as an enhanced physical training environment. The first example, from the 1980s, involved juggling. If virtual juggling balls fly more slowly than real balls, then they are easier to juggle. You can then gradually increase the speed, in order to provide a more gradual path for improving skills than would be available in physical reality. (This idea came about initially because it was so hard to make early VR systems go as fast as the reality they were emulating. In the old VPL Research lab, where a lot of VR tools were initially prototyped, we were motivated to be alert for potential virtues hiding within the limitations of the era.) Variations on this strategy have become well established. For instance, patients are learning to use prosthetic limbs more quickly by using VR these days.
Beyond rational argument, I was biased in other ways: The therapeutic use of VR seemed "too cute," and sounded too much like a press release in waiting.
Well, I was wrong. PTSD treatment in VR is now a well-established field with its own conferences, journals publishing well-repeated results, and clinical practitioners. Sadly, the Iraq war has provided all too many patients, and has also motivated increased funding for research in this subfield of VR applications.
One of the reasons I was wrong is that I didn't see that the same tactic we used on juggling balls (of gradually adapting the content and design of a virtual world to the instantaneous state of the user/inhabitant) could be applied in a less somatic way. For instance, in some clinical protocols, a traumatic event is represented in VR with gradually changing levels of realism as part of the course of treatment.
Maybe I was locked into seeing VR through the filters of the limitations of its earliest years. Maybe I was too concerned about the cuteness factor. At any rate, I'm glad there was a diversity of mindsets in the research community so that others could see where I didn't.
I'm concerned that diversity of thought in some of the microclimates of the scientific community is narrowing these days instead of broadening. I blame the nature of certain online tools. Tools like the Wikipedia encourage the false worldview that we already know enough to agree on a single account of reality, and anonymous blog comment rolls can bring out mob-like behaviors in young scientists who use them.
At any rate, one of the consolations of science is that being wrong on
occasion lets you know you don't know everything and motivates renewed
curiosity. Being aware of being wrong once in a while keeps you young.
