Is AI the new light bulb?
Virtual reality (VR) and artificial intelligence (AI) remind me of stories I’ve read describing the late 1800s and early 1900s when people were finding novel uses for electricity, like electrocuting elephants and lighting up the streets at night.
Personally, I saw such follow-on developments technology in person with the minicomputer era, again in the PC revolution, and once more with the smartphone. I think that’s where we are today with AI, deep learning, and VR, and to a lesser extent AR.
Electricity was an expensive, privileged resource in the 1800s. It took a long time for it to become a commodity that was competitive and universally distributed in most parts of the world. Hotels used to have electric meters in them, charging for usage, just as some hotels do today for Wi-Fi. But with the advent of electric refrigerators, stoves, clothes irons, radios, TVs, mix masters and garage door openers, electricity became as expected and needed as water (which, once again, is becoming a valuable commodity, and easy access is reserved for the privileged.)
Likewise, computing was an exotic resource for the military industrial priesthood. It guided airplanes and rockets and banking. Corporations used computers for accounting and inventory control, and the government computerized the census. As the applications increased, so did sales, and that drove down prices and stimulated research leading to the PC and smartphone and compute power for all.
In 2016, we saw the explosion of interest, articles, and conferences (OMG, the conferences) in VR, and less so in AI.
Artificial intelligence has attracted a steady level of interest (as measured by Google trends), whereas the subset, deep learning (see article, Emerging Companies Summit, page 4) has been growing in interest. Interest in virtual reality was pretty low until early January 2015, nine months after the astonishing windfall investment Oculus got from Facebook, and then it exploded. We have found forty-eight (48!) VR conferences, almost one a week sometimes. That’s insane. The granddaddy of VR conferences is IEEE VR, which marked its 23rd anniversary this year. (A side note: Look at the list of topics at the early IEEE VR conferences and compare them to today; about 85% are the same subject. Conclusion—problem not solved.)
However, based on developer (if not consumer) enthusiasm, VR (to quote Roy Taylor of AMD) isn’t going to be a big thing, it’s going to be everything. Everything—like electricity is to us now—ubiquitous, cheap, and critical to our survival and well-being. Saying that stretches the imagination to envision clamping a box on your face for survival and well-being. However, some of the therapy treatments being developed with VR could indeed be so characterized.
It is more understandable that artificial intelligence will become a major component in our daily lives. For many of us it already is; case in point is the natural language applications and devices like Amazon Echo (Alexa), Google home, Microsoft’s Cortana, Apple’s Siri, and “Hey Google.”
One of my favorite truisms is the technology works when it disappears. One of the things that helped kill Google Glass was its obnoxious attention-demanding appearance—much like the brick-sized mobile phone with its whip antenna.
Cooper, in addition to being a brilliant engineer, was a visionary, perhaps inspired by the Star Trek Communicator from the 1964 TV series. But it took 34 years for Cooper’s vision to be realized (and 50 for Gene Roddenberry’s).
VR is in a similar position. It was never targeted at consumers when the concepts began to gel in the early 1990s; it was a tool for researchers. There was perhaps the hope by some of the visionaries that VR could become a consumer product and experience. Because it was, and still is, a scientific tool, scientists have had conferences about VR ever since. And if you look at the sales of VR HMDs, it’s the scientific, industrial, military, and research sectors that are the big buyers. If VR follows in the footsteps of other technologies, it will be 20 or more years before it is as useful, ubiquitous, and necessary as electricity.
AI, however, is already on its way, and it’s invisible, so therefore it must be working. Here’s where a bromide fails: I don’t know anything about AI, but I’ll know it when I see it. No you won’t— can you see electricity?