The question of nature
by Ian Simmons
[ people - october 02 ]
Jaron Lanier is a computer scientist, composer, artist, and author, best known for his work in virtual reality, a term he coined. He has performed with Philip Glass, Ornette Coleman, Vernon Reid and Terry Riley, among others, and writes chamber and orchestral music. His art has been exhibited in the US and Europe. He has written for the NY Times, Wall Street Journal, Forbes, Harpers Magazine, Wired Magazine and Scientific American, and he has been profiled on the front pages of the WSJ and the NY Times.
* * *
Ian Simmons (IS): Steven Spielberg picked your brains for Minority Report. Which bits of your input made it to the screen?
Jaron Lanier (JL): The information technology stuff - the computers, three-dimensional camcorders, communication tools, the advertising, the cereal box, most of which already exist. They are versions of stuff in my lab, some of it quite old. So the movie, which is set 50 years in the future, is more retro than futuristic and will be regarded as excessively conservative in 50 years.
IS: Seeing advertising reading one's eyeballs, it struck me that no one was wearing mirror shades...
JL: I brought in a demo of a facial recognition technology. (Eyematic, where I'm chief scientist, has the highest-performance face recognition.) I made a passing comment about biometrics in the future. The argument for using the face for identification is that there is back-up: humans are also good at it too. I made the point that as with any measure, there might be unexpected side-effects. If you used iris-scanning, would criminals start stealing eyeballs? I said it as an argument for why iris-scanning wouldn't be used in the future, but Spielberg, as I recall it, just seized upon it: "Eyeballs! We can do a lot with that!" It was an inadvertent inspiration, but it works well dramatically... The technology exists to tell if an eyeball is alive, but at our cameras are not capable of sufficiently high resolution. In 50 years, though, that will be trivially easy. A whole set of the issues portrayed in the movie are simply not going to happen.
IS: Were you a Philip K Dick fan before you were hauled into the movie?
JL: No. Last week I was invited to a party for science fiction writers, who were all profoundly offended that I was unaware of what had been written in the last year. I've read little recent science fiction other than the cyberpunk novelists, whose work is directly relevant to mine and who often make me a character in their novels - I make a point of reading novels in which I appear! William Gibson had me as a head floating in cyberspace, and Bruce Sterling and Neal Stephenson inserted me as cameos...
I read Philip K Dick as a kid, and I was interested in his particular dementia. He experienced himself as living in two time-frames - biblical and present-day - at once. There is an issue, for me, of what I would call the inappropriate hipness of paranoia. It's a legacy of the Sixties, and it probably relates to the fact that some of the drugs that were celebrated then have a side-effect of inducing a level of paranoia. There are still some circles in which you are not regarded as credible if you don't exhibit some level of paranoia. Philip K Dick was certainly not the only practitioner (Thomas Pynchon and many others fall into the category), but there is that legacy of fashionability. I think it stinks. Paranoia is a form of choosing victimhood over self-empowerment. There is nothing recommend it. The slogan is that it's not paranoia if they're really out to get you, but the point is that it is. What is true about the world is a separate question to your attitude towards it. If your attitude is that in order to be hip you have to believe that what's really going on is somehow inaccessible to you and involves people who are more powerful, clever and stealthy than you, you make yourself stupid. In truth, there isn't that much variety between human beings. The intellectual and bodily powers of a corporate head or a king, his emotional range, is not that different from yours, and there's no need at assign supernatural status to any other human being.
IS: So you subscribe to the cock-up rather than the conspiracy theory of history?
JL: My theory of history is fundamentally optimistic. Sometimes reality is genuinely tragic, and the project of trying to live together as individuals is genuinely hard. It is difficult to be the human race, we've done really well, and I discern a slight improvement over time which is evidence for that.
IS: What will be the key drivers for science, technology and life for the next 50 years?
JL: The principal one is very simple to identify: what is true about nature. There's so much we don't know about nature, about how complicated biology is: with just a few clever insights, the difficult questions - how the brain processes information, or how DNA represents shapes or organisms - might be solved more quickly and simply than we can imagine, but right now we don't even know how hard the questions are. I'm confident that within 50 years, if they're easy, we'll have answered them; and if they're hard, we'll have some idea about how hard they are.
IS: So, biology will be for the 21st century very much as physics was for the 20th century?
JL: No. With physics, we are usually under the impression that it's easy. Since pre-Newtonian times, since Kepler, we've believed that physics was reducible and all we had to do was find the right formula. A century ago, a lot of physicists believed that physics was completed; they didn't grasp that relativity and quantum mechanics were around the corner. We've never felt that physics was just too complicated, whereas biology seems to be irreducibly complex. Perhaps we'll be surprised in an symmetrical way and it will turn out to be easier than we think, but who knows?
After solving the nature question, the next driver will be how good we get with computers - not how fast we can make them, but a more subtle question of how well we can program them. Faster computers have enabled us to do more things, and yet the interesting question is not how far Moore's Law might go (and certainly there are debates about that), but whether we can learn to write giant programs. If you could beam a computer back from 50 years in the future, it might be a billion times faster, but what we could do with it would be restricted.
If it turns out that there is some incredibly simple way for biologists to explain what currently seems complicated, computers won't be so important. If, on the other hand, these questions remain unsolvable, our agility and ability with software will be the critical limiting factor. Really complicated programs don't travel. I've been intimately involved in surgical simulation. High-quality surgical simulation captures the grad students who initially worked on them and enslaves them for life! They become an irreplaceable portion of that program, which creates a structural limitation on how useful computers can be for large models. They interfere with the scientific method (in that you no longer have the ability to replicate) and they destroy the scientific community. The other problem is that we just can't write the big programs with 100m lines of code...
IS: What are you working on at the moment?
JL: I'm trying to reduce the amount of things I do. I ended up with a lot of different collaborations in different places and got too spread-out. I'm continuing research into telemergence, which involves creating presence at a distance both for the purpose of creating a simulated transportation booth and also to study cognition. By fully creating the illusion that another person is there in front of you, you produce the perfect tool for studying how people perceive each other. I'm also continuing the surgical simulation research, which is basically about making better models of the body so that we have a live model of a patient being operated on; eventually surgery will blend with new ideas such as stem-cell therapies.
I don't have a clever name for the stuff I'm working on now. The notion is to create a sort of computing which is not based on protocol. The idea of the protocol is that you have a sender and a receiver of information, each of which know the format of that information. 'Proto' means 'in advance'. All our computer theory and computer science is based on this idea, but it's not the only way to build complex systems. If you look at nature, there are some cases where the sender and the receiver agree the format in advance: an infant recognises the human face because of a prior agreement between the infant's brain and the DNA that structured the face; colour codes are shared between species. So there are protocols in natural communication, and even the term 'communication' suggests advance agreement.
We tend to describe information transfer in which there was no prior agreement as pattern recognition or response; we have a different set of terms for it. Whatever we call it, there are many examples in nature: our nervous system learns to apprehend many things it hasn't seen before for which there was no prior agreement. There is a different error profile. Pattern recognition involves a gradual give-and-take, and errors are likely to be gradual rather than instant and total. It's that feature of recognition that interests me. I'm attempting to build computer architectures and programming languages in which pieces recognise each other rather than adhering to a protocol.
IS: That sounds like it might go some way towards solving the question of what a really big computer is going to do...
JL: That's precisely my hope. I also believe there is pretty strong circumstantial evidence that this kind of connection between elements is essential to the embryogenesis of the brain. There may be a 'starter' protocol, but parts of the brain learn to recognise each other as a child grows up and, furthermore, continue to do so over time. We've tended to base all our neuroscience on the same protocol metaphor which has dominated information science, and I think it could be completely wrong. This my creative passion at the moment...
IS:... and probably the thing most likely to make a huge difference. You really got on the map in the late Eighties with the Dataglove and related stuff.
JL: Tom Cruise wears one in Minority Report, but we can do all that stuff using machine vision. The first Dataglove was in 1979. There was tremendous hype when it came out - I hyped it just to get funded, but then folk went mad.
I think of it in terms of early childhood psychology. If you're small enough, you can't separate reality from fantasy. You're more powerful in fantasy, you're Superman; think of something and it pops into being. In reality, you're slower than adults, and really weak. It's a drag. Every child has a fantasy of a shared world that's also under the control of imagination.
The notion was to have an externalised, genuinely shared objective version of imagination and dreaming, and it was this primal desire that drove the frenzy. There was a lot of really egregious virtual reality being developed back then. I had no business starting a company that sold hardware. We were amateurs who suddenly became manufacturers. The problem with our stuff was reliability.
IS: You have a double life as a musician. What do you play?
JL: I have a bit of an addiction problem with instruments - It's deeper and healthier than heroin, I guess!
IS: I was reading an interview in Wire with Tom Waits, who has a similar addiction and has just bought a Calliope...
JL: His addiction is different. He tends to collect oddball, primitive instruments, American folk instruments. I have a tendency to think of instruments as somatic time-travel machines. If you learn to play an instrument from some other culture, you have a means to access that culture that no other artefact offers. You have to learn to move your body in the same way as they did. I think of an instrument as a sort of cam that records the negative space of the motions and methods of playing that some other culture preferred, and you enter into their universe of body motion. I find that to be a very interesting experience. There really is a whole internal peri-perceptive culture to each period and place, and musical instruments are our only record of it. And that relate directly to my interest in virtual reality - virtual spaces, how we perceive where we are, and how we set up a cognitive strategy for modelling our world and interacting with it. Instruments are the only historical record of how people have done that. The quality of a lot of contemporary music is extraordinary. What is curious to me - puzzling, let's say - is that this is the first period since the dawn of recording when there hasn't been a new style in a long time. We seem to have reached a stage of what anthropologists call "pattern exhaustion" which probably started in the Eighties. Grunge was marketed as new, whereas, in fact, it is very like Sixties music and was more retro than new. The difference between punk and hippie rock - or hippie rock and big-band music - was extreme. Right now you have to be very involved in music to appreciate the differences. Stylistic evolution has stopped: why? There are different theories. Maybe culture is going cyber, abandoning Gutenberg and adopting the Internet; human communication is no longer something that gets lost but is saved in this enormous shared brain where everything is recycled. That's the optimistic way of looking at it.
When I was a teenager, the dynamic was that elders were supposed to be shocked at the level of obscenity that youth displayed. I suppose I am now an elder, but my concern is almost the opposite: not that they aren't going far enough with obscenity (they're doing just fine on that front!), more that they haven't experienced a level of intensity and joy in music. They buy a lot of music, despite Napster, but it's an accessory, not an overwhelming experience. It's now more casual, which is a curiosity for me.
As for what I listen to, I would plug my own stuff, except I got bored with the recording industry, and it's difficult to get it now. A lot of the cultural cross-over stuff is interesting: George Brooks, Jay Uttal, a guy in the San Francisco area whose music is a cross of Indian music and jazz which achieves results which I find fascinating. And a Persian-American woman I play with a lot, Susan Dehim. What happens a lot in the US is that radio stations that play mainstream hip-hop and r 'n' b during the day will let people come in at one or two in the morning and do live remixing. Some of that stuff is wonderful. If we believe that the lack of style is about a shift to recirculation, then the cool thing would be the mash-up movement, some of which is witty and wonderful.
Despite my reservations about mainstream pop, a lot of it is excellent, though one of the things about music being impermanent is that artists go by so fast...
IS: In Britain we're inflicted with bands which are reconstituting 1972 but not as well...
JL: Just be grateful you're not in Italy or France, where the pop music is astonishingly bad! I had an idea of doing an annual compilation of the year's worst Euro-pop as a camp thing.
There are interesting things going on almost unnoticed in the spiritual/ minimalist area of classical music. An interesting economics issue comes up: we have a world-wide society in the web of industrialised countries, with an enormous number of reasonably well educated, well-off people. A huge number of people are producing music which, 50 years ago, would have been earth-shaking, but because there is so much of it, we can't know about it. The ratio of creators to consumers has shifted, so there are far fewer people who would consider themselves as consumers. Which is, of course, a good thing, but... Historically, a lot of the great musics of the world were created by people who had to play to make a living. Now, even in traditional cultures, you have people with passion who have to figure out to make a living despite music. That reversal is profound, and it's a structural problem that will have to be reckoned with.
IS: How did you take these varied paths?
JL: I wouldn't know where to begin. I've been drawn to trying to find out how people connect. That may have something to do with losing my mother when I was very young. The fact that communication is possible between people seems to be the greatest, most mysterious phenomenon, far greater than quantum gravity. I've also been drawn to what makes an experience exotic, the game theory of novelty, if you like. Those two themes are ever-present in everything I've done. But there are limitation as to how well you can analyse yourself.
