Join the conversation at CRNtalk!
CONDUCTED AUGUST 2001
Question 1: Tell us a little about yourself. How long have you been interested in emerging technologies? How long have you been committed to informing the public of future changes?
Interested? Since I was a 12 year old, 45 years ago. I started writing about these topics at university, nearly four decades ago.
Question 2: When and under what circumstances did you first become aware of the concept of the Spike? Has your concept of The Spike evolved since you first became aware of the concept?
As I mention in the book, the idea of runaway self-bootstrapping change is something I first met in an Analog article by Harry Stine back in 1961. He claimed that certain exponential factors in engineering would go vertical by the mid-'80s. Each human would have the energy of star at his disposal (of course it was "his" in those days). Didn't happen. Later, Vernor Vinge proposed a technological singularity in Marooned in Real Time, and said he was quite serious about it. I wondered if he might be right.
Question 3: Many people are skeptical of the concept of a Spike because it seems fundamentally counterintuitive. For instance, planes, automobiles, mechanical tools, and buildings don't seem to have changed much during the past couple of decades. What is your response to such skepticism?
Very sensible to
be skeptical, but not for that reason. In fact, cars have changed rather
a lot; Tin Lizzies didn't have little computers
Question 4: How does your concept of a Spike differ from the predictions of Vernor Vinge and Ray Kurzweil? Have you discussed the subject with either of them?
Dr Vinge is more cautious than I am, but then he's a scientist, until recently an academic, so to some extent has to watch his back (and also he knows a great deal more about computers than I do). But other well-placed folks who've investigated the topic arrive at even more radical extrapolations than his, and I detail some of these in my book. I haven't met Ray Kurzweil yet, but his book The Age of Spiritual Machines and his forthcoming The Singularity is Near agree in large measure with my general sense of the thing. Maybe it's worth mentioning that I got there before Kurzweil; the original edition of The Spike came out in Australia in 1997.
Question 5: Many workers in the field of Artificial Intelligence and Molecular Nanotechnology feel hampered by research and development budgets that are overwhelmingly focused on the short term. Could such short term fixations derail AI progress and/or prevent a Spike?
In the short term.
Question 6: In your book The Spike, you seem unsure about the long term prospects for Drexlerian Molecular Nanotechnology. Would it be fair to say that you are a skeptic of Drexlerian MNT?
In the September 2001 Scientific American, Dr Drexler claimed that none of his critics had managed to find crucial holes in his engineering, laid out in great detail in 1992's Nanosystems. Other expert papers in the same issue acted as if machine-phase nanotechnology were impossible, or at least incredibly remote. I dunno, mate. There are some smart people working in this area, and the payoff would be enormous. It might be that the first fruits will come instead from proteomics and other biology-based sources of tiny useful gadgets. The key idea is surely on the money, though: compiling manufactured goods cheaply from the molecular scale up, rather than carving lumps of stuff expensively down to size.
Question 7: Some AI researchers such as Marvin Minsky argue that the question of AI revolves primarily around software -- Minsky has stated that a 1 megahertz machine could become sentient with the right programming. Other researchers, such as Hans Moravec, seem to argue the opposite, that hardware is the main limiting feature. What is your opinion?
Both, wouldn't you think? A human brain embodies a lot of software in its wetware or firmware or whatever you want to call it, and a lot more gets programmed in from a structured environment, and develops further in run-time. The surprising answer might be: first, get a huge amount of raw grunt (available, all things being equal, inside 15 or 20 years), then load in a nano-scale scanned brain. Maybe a chimp, or a dying paraplegic person might volunteer. See how the code runs. You'd need to give the system a lot of high-grade sensors and effectors.
Question 8: How long do you believe Moore's law will continue? Pessimists argue that it will end in the next couple of years because of exploding fab costs. Optimists maintain that it will continue for decades. If the pessimists are correct, could there still be a Spike?
Everything is connected. Silicon isn't the only fruit. The other day I saw news from IBM about a single molecule carbon NOT gate. Dear god.
Not a chip coming to a store near you any day soon, but it changes the rules, doesn't it? Except that the meta rule says: expect something to jump out of the box in time to keep the second order Moore's `Law' curve running up the graph.
Question 9: Writers in the 1960s argued that the main challenges of the year 2001 would be exploring and colonizing space and keeping people who only worked 15 hours a week occupied. You discuss the history of off-target predictions in your book The Spike. Does the long history of inaccurate predictions give you cause for pause?
Of course. They were all so boringly unimaginative. It takes a huge leap to get the future even remotely right. That's one reason I explore several delightful but apparently off-the-wall ideas in the book: J. Storrs Hall's idea of `Utility Fog', for example, a pervasive landscape of nano gadgets suffusing the air you breathe, that can be morphed into all kinds of usefully goodies. Probably too conservative by half.
Question 10: Will the current economic downturn in the computer industry have long term negative effects, delaying or even preventing a Spike? How can semiconductor companies such as Intel continue to fund R&D if revenues stay low?
Question 11: Do you believe that future technological advances will come more from "wet" technology - biogenetics and cell research, or "dry" technology -- microchips and fiber optics? Will biologists and chemists or mechanical and electrical engineers play a bigger role?
As I indicated
above, both have a role to play. Genome research is pushing the development of
humungous supercomputers to analyze protein folding, the next big frontier. That
might feed back into machines built on those very principles. Meanwhile, quantum
computing might emerge big time.
Question 12: What are your plans for the future?
Hang on as long as
possible, waiting for the cornucopia machines. Oh, and I want my hair back. :)
This interview was conducted by Sander Olson. The opinions expressed do not necessarily represent those of CRN.
Copyright © 2002-2008 Center for Responsible Nanotechnology TM CRN was an affiliate of World Care®, an international, non-profit, 501(c)(3) organization.