Dr. Frank takes us on a journey from the evolution of technology into the distant future, and talks about what life will be like in dozens, hundreds, or billions of years. Is energy finite? Will technology become stagnant? What can we do? Is there hope? He takes us through all these questions and brings us back safely. Good food for thought.
Today I'd like to wax philosophical about a fun, if somewhat off-the-wall topic. Every now and then, some futurist or visionary will write a book in which they try to project the evolution of technology into the distant future, and talk about what life will be like in dozens, hundreds, or billions of years. Clearly, any such discussion involves a lot of speculation, but it can still be refreshing to step back every now and then from the day-to-day grind of engineering work, and think about the possible long-term implications of what we are all doing.
For example, Ray Kurzweil observed in his book The Age of Spiritual Machines that the trend of exponentially improving logic cost-performance in the IC era can be viewed as only the most recent phase of a much longer and super-exponential trend that spans multiple technologies, from hand-cranked mechanical desk calculators, through relays, vacuum tubes, discrete transistors, and finally ICs. Ray argues that positive feedback loops underlying the social process of technology development are responsible for this seemingly technology-independent long-term trend. Even more broadly, he ties this observation into the context of the ever-accelerating development of human civilization, intelligence, and life, all the way back to the first primordial ooze.
Our present era, in which MOSFET leakage problems and other more fundamental physical limits threaten to (perhaps permanently) cause our progress to plateau even before the end of the ITRS roadmap, may be unique, in the sense that, if the physical limits that we can see looming on the horizon today are indeed truly insurmountable, then this would be the first moment in the 4.5 billion year history of life on Earth that the evolution of the planet's information processing "technology" (construed broadly, including "innovations" like DNA and the human brain) is about to stop its exponentially accelerating pace, and instead settle into some sort of permanent equilibrium. And who knows, perhaps this will indeed happen—we can't rule it out. But it is interesting to note that it would be completely unprecedented in our history.
However, if a permanent (as opposed to temporary) stagnation of our technology does occur, it would have rather dire long-term consequences. In a variety of articles over the last five years (most recently, this one), astrophysicists Lawrence Krauss and Glenn Starkman have discussed the long-term prospects for life in the universe. Using the latest observations of the accelerating cosmological expansion, they argue that the total amount of energy that life can ever harness in the universe is finite—eventually, all but a handful of galaxies will be receding from us so rapidly that we could never catch up with them even at the speed of light. Thus, we can think of the part of the universe that is accessible to our descendants as being a sort of vast cosmic battery, one that will eventually run out, spelling the end of all life.
Or, will it? Krauss and Starkman's argument appears to depend crucially on their assumption that a fixed amount of energy is required to perform any "interesting" operation, that is, one that could be a meaningful step in any ongoing process that merits being called life, or computation. On the other hand, suppose that it is possible, before half the available energy runs out, that we can construct a new form of intelligent life (perhaps artificial life in a computer) that is twice as energy efficient as the previous life, that is, it uses half as much energy to complete the same total number of interesting computational operations or "thoughts." And then, suppose this progeny invents a new form of life that is twice as energy-efficient yet. And so forth. Like Zeno's hare which can never catch up with the tortoise, in principle, an infinite number of steps of computation (or of "living") could occur while never fully depleting the finite supply of energy. That is, this is mathematically possible, assuming that there is no fixed lower bound on the energy consumed per operation.
Fortunately, as far as we know today, there is no such bound. The widely known von Neumann-Landauer limit of kT ln 2 only applies to logically irreversible operations, and Charles Bennett of IBM showed in 1973 that an indefinitely large amount of interesting reversible computational work can be done, at least in principle, without falling prey to this limit. If reversible computing can indeed be successfully implemented, and if it can be made ever more energy-efficient over time, then at least there is hope that some (computational) form of life might continue in this universe literally forever. But if, on the other hand, reversible computing is impossible, as many have claimed (but not rigorously proven), then Krauss and Starkman are right, and all life in the universe is eventually doomed to fade away, although it could take many billions of years.
One may think that it is much too early to worry about such issues, but in a sense, it may be almost too late already. And that is because, supposing that we were to soon lose the technological momentum that propelled progress forwards at a relentlessly faster pace throughout the past century, is there any guarantee that the acceleration would ever start up again? Perhaps it would But still, sometimes I worry that, for all we know, we could be at a pivotal juncture right now in the history of life in the universe, and that we might not be trying quite hard enough to work towards the critical breakthroughs, such as reversible computing, that will be needed, in the long run, to save literally everything.
This thought is often what jolts me out of bed in the morning. Of course, one man can't accomplish very much all by himself. I can only hope to find, and inspire, a few others who will pick up the torch and run with it. And that they, in turn, find a few others And so on, until we have reached a "critical mass" of engineers, physicists, and deal-makers who understand well what needs to be done. Despite the astrophysicists' gloom, an eternal future for our posterity may still be possible, for all we know for sure today. If we try, we may side-step a purgatory of technological stagnation leading to eventual decline. The infinite future may yet be saved, via sufficiently aggressive power management, based on new reversible computing breakthroughs. But, we have to really want it.
We now return you to your regularly scheduled workday
Dr. Michael P. Frank, Assistant Professor
FAMU-FSU College of Engineering
Department of Electrical & Computer Engineering
2525 Pottsdamer St, Rm 341, Tallahassee FL 32310
firstname.lastname@example.org, phone 850-410-6463, cell 597-2046