I attended Douglas Hofstadter's Symposium titled
"Will Spiritual Robots Replace
Humanity by 2100?" on April 1, 2000 at Stanford University. Among other
things, I was hoping to find out what the experts in the fields of
nanotechnology and robotics had to say about this topic of discussion. Is
there really a chance that robots will be thinking and acting on their own
behalf by 2100? And if so, will they be replacing us -- either with or
without our consent and cooperation?
Douglas and his eight guests talked for five hours about their vision of humanity's future... as each panelist looked through a telescope with the lenses of his own particular area of expertise into the future. Many speakers cited Gordon Moore's Law of the ever-increasing pace of technological changes to make the point that technology is changing faster than ever before, and that rate of change is expected to increase at an exponential rate -- so it is very difficult to predict where we will be in one hundred years from now. Douglas explained that he only invited guests who agreed that there is a possibility for robots to be spiritual. Douglas wanted to focus on the question of "Who will be we in 2093?", since a visualization of who we will be is at the core of how we can understand how we might be utilizing new technologies.
The most surprising and noteworthy speech was given by SUN microsystem's co-founder Bill Joy. Joy presented an impassioned plea that we relinquish and renounce further development of three potentially dangerous technologies -- genetic engineering, nanotechnology, and robotics. Joy's arguments in favor of relinquishment appear in an article in "Wired" magazine this March. Joy explained that mankind has reached a point where we can choose to act responsibly... or we can continue seeking truth and knowledge for its own sake. Joy recommends that we stop and recognize the dangers of democratizing the availability of self-replicating (ie: potentially out-of-control) technologies, which could conceivably lead to pestilences the likes of which we haven't seen for hundreds of years. Unchecked use of self-replicating technologies can lead to genocide or extinction, and Knowledge Enabled Mass Destruction (KEMD) is a risk for us when knowledge itself becomes a weapon. In the 20th century, the nuclear, biological and chemical weapons were our biggest threats -- yet these were all controlled by the defense departments of a handful of countries, and had limited commercial value and public access. In the 21st century, Joy anticipates that our biggest threats will be from genetic engineering, nanotechnology, and robotics. The danger of these new technologies is that they will be designed to be self-replicating and ever-evolving by nature, they will be developed primarily by the commercial sector, and they will not require rare raw materials. Bill Joy feels the only viable solution to what our 30%-50% chance of extinction from new technologies is voluntary relinquishment of these technologies, just as we renounced development of biological weapons a few decades ago.
As if in silent support of Bill Joy's speech, numerous glitches in the lighting, remote microphones, and slides frequently punctuated this symposium.... reminding all present that technology as we know it today is far from fool-proof and 100% under control.
Other panelists did not join Bill Joy's bandwagon, and nanotechnologist Ralph Merkle in particular expressed strong feelings of disagreement. Merkle pointed out that if we relinquish nanotechnology, less scrupulous nations and individuals (such as Saddam Hussein) will develop it first. Ralph said that "machines people make bear little resemblance to living systems", and pointed out how ludicrous the idea of robotic machines running amok sounds. He said that "a feral 747" sounds incongruous. Merkle then explained that broadcast architecture can be an effective safeguard against out-of-control nanotechnology, since blueprints are not located in the individual units but are instead broadcast from a central point which can regulate replication more simply and with better controls. Merkle believes we can design intrinsically safe self-replicating systems. Merkle also believes we need to further study new technologies to determine whether they might be used for defense (such as castles) or offense (such as nuclear weapons) in order to better evaluate the risk of deliberate abuse and accidents.
Mobile robot research pioneer Hans Moravec believes that humans are an interpretation of the pattern, and that we can model a person as an I/O device... like a wind-up toy. An ET might in fact see us this way. Robots will naturally get that interpretation, because we'll see them that way. Robots will have thoughts and feelings because we will tell them of these things, and they will INSIST they have feelings and beliefs, too. Moravec believes, "Physical reality itself is just a way of looking at things. We don't really exist. We only think we do." Regarding the possibility of deranged robots running amok, Hans responded, "I worry about robots getting strange beliefs by accident. I suggest there should be a policing system among themselves. I believe it's fully possible for machines to go crazy... then you have police."
Ray Kurzweil (author of THE AGE OF SPIRITUAL MACHINES) believes that we can't go back to less technology, but "we can find sophisticated ways to manage the double-edged sword of technology". Ray believes that what's supremely spiritual is our pattern, not our biology. Even the particles in our biological bodies are currently replaced in a period of months, yet we consider ourselves to still be the same individual. In the future, Kurzweil envisions that nanobots will lead to virtual reality with full sensory perception, and these bodies we have in VR can be anything we want them to be at any time. Kurzweil said that he uses the term spiritual to really mean conscious... to have feelings and be aware of them. Ray thinks humanity will exist in 2100 if we address the dangers Bill Joy mentions; "the future's not yet written".
John Holland, the inventor of genetic algorithms, pointed out that "we need a decent theory", because "theory is the sine qua non for progress in the Complex Adaptive System (CAS) domain". Examples of complex adaptive systems are the Central Nervous Systems (CNS) in our bodies, morphogenetic processes, and ecosystems. While it's true that evolutionary systems can solve well-defined problems, they are not very useful for more open-ended problems like "How do I optimize an ecosystem?" Holland pointed out the need for better theory and different techniques in order to move from the realm of raster-scanned input, switching see-this-input-make-that-output operations, and Boolean logic and into the more highly entangled realm of visual input that saccades (jumps to salient points), reverberating modulation operations of ongoing semi-autonomous activity, and analogy. While Holland agrees with Bill Joy's theory that we need to relinquish some technologies, he clarifies that we need to prune the bad paths in order to avoid making the BIG mistakes.
Kevin Kelly, author of OUT OF CONTROL and editor at Wired magazine, says "We're headed for a 100-year identity crisis. And we're launching right now. Robots are mind-children... they are children of our mind. We should aim to train them to be good citizens... then let go. The power of technology is proportional to its apparent out-of-controlness." Kelly also stated that AI is a different name for ET... and professed his opinion that the ultimate spiritual experience is meeting an ET, which we will soon be encountering when Artificial Intelligence becomes a commonplace reality.
Frank Drake, head of the Search for Extra-Terrestrial Intelligence (SETI) Institute, described how SETI could illuminate many things for us if we discover an extra-terrestrial civilization. Such a civilization is likely to be hundreds or thousands of years ahead of us, giving us guidance for answering the question, "Who will we be?" On a cosmic and biological time scale, we'll have computing capacity far beyond what we've had which will allow us to plan our searches for extra-terrestrial intelligence better. We are approaching the 40th anniversary of the first search (conducted 4/8/1960), and we've seen technological improvements that have out-stripped Moore's Law in that time.
John Koza, the inventor of genetic programming in AI, believes that we have a long way to go before we have spiritual machines, which will require neural implants, sensory enhancement, and neural scans (where brain activity can be downloaded to silicone). Koza stated, "The time scale is WAY wrong" for spiritual robots by 2100, and explained how one of our brain seconds (BS) is equivalent to ten to the fifteenth operations... and we have 10 to the twelfth power neurons in our brains. By the year 2100, then, "we're not talking about consciousness or the Turing test... we're talking about Turing Lite."