In our September issue, Michael Dertouzos wrote a column, “Not by Reason Alone,” that took Bill Joy of Sun Microsystems to task for a piece Joy had written in Wired. In his Wired article, Joy argued that humanity should renounce certain lines of research, including nanotechnology, because of the dangers they pose. Dertouzos argued that Joy’s view was flawed because his predictions were based on reason which, taken alone, is an inadequate guide to the future. Dertouzos’ column drew an impassioned response from Ray Kurzweil, author of The Age of Spiritual Machines. We print Kurzweil’s letter and Dertouzos’ rejoinder.
Although i agree with michael dertouzos’ conclusion in rejecting Bill Joy’s prescription to relinquish “our pursuit of certain kinds of knowledge,” I come to this view through a very different route. Although I am often paired with Bill Joy as the technology optimist versus Bill’s pessimism, I do share his concerns about the dangers of self-replicating technologies. Michael is being shortsighted in his skepticism.
Michael writes that “just because chips…are getting faster doesn’t mean they’ll get smarter, let alone lead to self-replication.” First of all, machines are already “getting smarter.” As just one of many contemporary examples, I’ve recently held conversations with a person who speaks only German by translating my English speech in real time into human-sounding German speech (by combining speech recognition, language translation and speech synthesis) and similarly converting their spoken German replies into English speech. Although not perfect, this capability was not feasible at all just a few years ago. The intelligence of our technology does not need to be at human levels to be dangerous. Second, the implication that self-replication is harder than intelligence is not accurate. Software viruses, although not very intelligent, are self-replicating as well as being potentially destructive. Bioengineered biological viruses are not far behind. As for nanotechnology-based self-replication, that’s further out, but the consensus in that community is this will be feasible in the 2020s, if not sooner.
Many long-range forecasts of technical feasibility in future time periods dramatically underestimate the power of future technology because they are based on what I call the “intuitive linear” view of technological progress rather than the “historical exponential” view. When people think of a future period, they intuitively assume that the current rate of progress will continue for the period being considered. However, careful consideration of the pace of technology shows that the rate of progress is not constant, but it is human nature to adapt to the changing pace, so the intuitive view is that the pace will continue at the current rate. It is typical, therefore, that even sophisticated commentators, when considering the future, extrapolate the current pace of change over the next 10 years or 100 years to determine their expectations. This is why I call this way of looking at the future the “intuitive linear” view.
But any serious consideration of the history of technology shows that technological change is at least exponential, not linear. There are a great many examples of this, including exponential trends in computation, communication, brain scanning, miniaturization and multiple aspects of biotechnology. One can examine this data in many different ways, on many different time scales and for a wide variety of different phenomena, and we find (at least) double exponential growth, a phenomenon I call the “law of accelerating returns.” The law of accelerating returns does not rely on an assumption of the continuation of Moore’s law, but is based on a rich model of diverse technological processes. What it clearly shows is that technology, particularly the pace of technological change, advances (at least) exponentially, not linearly, and has been doing so since the advent of technology. That is why people tend to overestimate what can be achieved in the short term (because we tend to leave out necessary details) but underestimate what can be achieved in the long term (because exponential growth is ignored).
This observation also applies to paradigm shift rates, which are currently doubling (approximately) every decade. So the technological progress in the 21st century will be equivalent to what would require (in the linear view) on the order of 20,000 years.