Borderlands of Science

by Charles Sheffield

  The big problem with SETI was stated with admirable succinctness by the great Italian scientist, Enrico Fermi: Where are they? If there are extraterrestrial intelligences, some are presumably more advanced than we are. Why haven't they showed up and presented themselves at the United Nations, or sent a proof of their existence that is impossible to miss or deny?

  This absence of contact is known as the Fermi Paradox, though it is hardly a paradox. It is simply a good question, to which there is no good answer. Some people suggest that we have not yet found the right radio frequency, or have looked in the wrong direction. Some argue that we are still in too primitive a state of technology, so that our proposed methods of sending or receiving signals are little better than the multicrop agricultural fields proposed by Gauss. And of course there are others who say that aliens don't need to signal, because they visit Earth on a daily basis in UFOs.

  But then there is the alternative viewpoint: We are alone, the only intelligent species in our galaxy. It is a waste of time and money scanning the sky for messages, or sending them out to nowhere.

  Is SETI a waste of time, as its critics say, because the probability of success is low? Or is it, as its disciples claim, a project that we ought to be engaged in all the time and at an increased level of effort, because the payoff of success could be so enormous?

  Although the United States Senate cut off all SETI funding in 1993, the effort continues with private support. The program that used to be at NASA Ames has moved, almost in its entirety, to the SETI Institute in Mountain View, CA. The Planetary Society, in Pasadena, continues an active search under Paul Horowitz at Harvard. A very readable background discussion of the whole subject can be found in The Search for Extraterrestrial Intelligence: Listening for Life in the Cosmos (McDonough, 1987).

  In science fiction, SETI has long been an accepted element of the field. Three good and very different examples are The Hercules Text (McDevitt, 1986); Contact (Sagan, 1985); and The Ophiuchi Hotline (Varley, 1977).

  9.4 Beam me up. We know how to send signals to other civilizations at the speed of light, and we are already looking for messages from them. But do we really need to go in person? Why not transmit a complete signal that represents you or me, and use it to re-create us at the other end? That way, we'll get to the stars as fast as possible, and so far as subjective experience is concerned it will be no time at all.

  Put aside for one moment the fact that there is nothing at the other end to put us back together. Ignore also the awkward question of which one is the real you—the one who was scanned back here, or the one who is reconstructed out there. Let us size the problem.

  The human body contains about 1028 atoms. To specify the substance of each atom (i.e. the element) needs only two decimal digits, since all atoms of a particular element are identical. However, we also need to specify information where each atom is. That calls for three coordinates, each given to an accuracy of, say, 2x10-10 meters, and with a maximum value of a couple of meters (basketball players have to crouch, or stay home).

  Associated with each coordinate we specify a number from 1 to 99 (for the appropriate element) with a zero when a coordinate lies outside your body. A representation of a complete human, down to hair part, birthmarks, and eye color, thus calls for about 1032 separate pieces of information. We assume that we have an understanding, in advance, that the (x,y,z) coordinates of atoms will be given sequentially, in a particular order.

  Let's see how long it will take to transmit a person from one place to another (distance is not relevant). A high-speed data link from ground to space is a few hundred million bits per second. We will be generous, and say we have a data link of a billion decimal digits (109) per second. Then the transfer of one human will take 1023 seconds, or 1015 years. The universe is only about 1010 years old. We would be better off using the Post Office.

  All right, we will seek economies. First, is it really necessary to send an exact description of every atom? As we know from heart, kidney, and liver transplants, these organs are all functionally similar. Let us send just the information defining the "real you," the brain and maybe a few glands that seem to define our emotions. We do not save very much. The range of each coordinate reduces from 2 meters to 20 centimeters. The transmission time comes down to about 1012 years. Still no good.

  How about if we simply regard the brain as a computer, and download the information held in it? This is certainly a popular science fiction device, although Penrose, as we will discuss in some detail in Chapter 13, would argue that it is impossible because the brain is more than a computer.

  Let us assume that he is wrong. There are about 1011 neurons in the brain. We number them sequentially, 1, 2, 3 . . . 1011. We make the (disputed) assumption that a neuron is a simple on-off device, so that its information can be represented by a single binary digit. Further, we assume that each neuron connects to an average of 50 other neurons (this is only an average; certain neurons in the cerebellum have up to 80,000 connections). Now we assume that the brain is completely defined by the neuron contents, plus all the neuron-to-neuron connections. For every neuron, we need to specify a binary digit, plus 50 decimal numbers each of which may be up to 11 digits long. Then a human, regarded solely as information, is defined totally by 1014 decimal digits. The transmission time using our billion-digit-a-second transmission system is a little over a day.

  This is an acceptable period. Even if we send the signal with triple redundancy, to make sure that the you-that-arrives is not subtly different from the you-that-was-sent, we are talking transmission times of a few days.

  Note that you will not exist physically until you have been downloaded from signal form into a clone of your body. That will be grown from your unique DNA description, which requires only about ten billion binary digits (a small fraction of the total signal) and could be sent as a lead file to the main message.

  What will we do when it turns out that the SETI signal is not the Encyclopedia Galactica at all, but the exact prescription for some alien interstellar tourist?

  9.5 A helping hand from relativity. It is Einstein's special theory of relativity that tells us we can never accelerate any object to move faster than the speed of light. The same theory, curiously enough, offers a helping hand when we want to travel long distances.

  As we mentioned in Chapter 2, one standard and experimentally tested consequence of relativity is time dilation. Let us recap its effects. When an object (in our case, a spacecraft) moves at close to the speed of light, time as measured onboard the spacecraft feels the same to the passengers, but as far as an external observer is concerned, it is slowed.

  The rule is very simple: for an object traveling at a fraction F of the speed of light, when an interval T passes in the rest of the universe, an interval only (1-F2) of T passes in the object's reference frame.

  Thus, if a ship travels at 90 percent of light-speed, time onboard relative to the outside universe is slowed by a factor 0.43; when a century passes on Earth, only 43 years pass on the ship. At 99 percent of light-speed, 14 years pass on the ship; at 99.9 percent of light-speed, 4 years pass on board. Clearly, if we can accelerate the ship close enough to light-speed—no mean feat, as we have seen already—then so far as the passengers are concerned, travel time to the stars or even to remote parts of the galaxy can be made tolerable. As an example, the center of the galaxy is about 30,000 light-years away. For a ship that traveled just one hundred meters a second slower than the speed of light, the perceived travel time from here to the galactic center would be only 24 years.

  Frank Tipler, grandly dismissing the practical details of ship drive design, has examined travel times for a ship that moves not at constant speed, but at constant acceleration (Tipler, 1996). Setting that acceleration at a comfortable one gee, Tipler finds that a round trip to the center of our galaxy will take about 40 years of shipboard time. The Andromeda Galaxy is about 2.2 million light-years away. A visit to it needs 57 years of ship time. And if we want to take a
longer trip, to the Virgo Cluster at 60,000,000 light-years distance, we can expect to be away for about 70 years. As we see, a constant acceleration telescopes almost all distances down to the point where a human lifetime is enough to travel them.

  The snag, of course, is that you might go to the center of the galaxy and back in one lifetime; but while you were gone, things here on Earth could be expected to change considerably in your 60,000-year absence. As for a trip to the Virgo Cluster, you could have left Earth when dinosaurs were the dominant land animals, and not be back yet.

  9.6 Faster than light. Like it or not, we have to explore the possibilities of faster-than-light travel. Without it, all our interstellar empires and intergalactic trade shows are impossible. What's the point of sending your army to quell an uprising when it happened 50 centuries ago, or ordering a piece of furniture that will take a thousand years to be delivered?

  We need a loophole. One possibility was suggested in Chapter 2, where the idea of quantum teleportation was explored. To find another one, let us return to Einstein. The assumption that we cannot travel faster than light is usually stated as one of the central elements of the theory of special relativity. In fact, what Einstein said was not quite that. You cannot accelerate an object faster than light, or even as fast as light. As you try to move something faster and faster, the energy needed to do it becomes greater and greater.

  However, this does not mean that particles which travel faster than light cannot exist. As one researcher into faster-than-light particles pointed out, that would be like saying that there can be no people north of the Himalayas, since no one can climb over the mountain ranges.

  In this case, the mountain range is the speed of light. Although no particle can be accelerated through the barrier, this in no way proves that particles cannot exist on the other side of the barrier.

  In 1967, Gerald Feinberg gave a name to hypothetical faster-than-light particles. He called them tachyons, from the Greek word tachys, meaning swift. Richard Tolman, as early as 1917, thought he had proved that the existence of tachyons would allow information transfer to the past, and thus allow history to be changed. For example, a message back to 1963 could in principle have prevented the Kennedy assassination. That possible use of tachyons was explored in the novel Timescape (Benford, 1979). Today, however, Tolman's argument is no longer accepted; Benford's novel remains as fiction.

  Tachyons do appear to be permissible within the framework of conventional physics, in that there seems to be no physical or logical law ruling out the possibility of their existence. This has led some writers to argue that tachyons must exist, adopting the rule of the anthill from The Once and Future King (White, 1958): "Everything not forbidden is compulsory."

  Suppose for the moment that tachyons are real. Then to see how light speed forms a natural barrier separating bradyons (the familiar "slow" particles of our universe, also known as tardyons) from tachyons, imagine that we accelerate a charged particle faster and faster, for example using an electromagnetic field.

  What happens to it? The particle certainly continues to increase in speed, but according to the theory of special relativity as it gets closer to the speed of light it also becomes more massive. As a result it becomes more difficult to accelerate. The mass doubles, then quadruples, and more and more energy is needed to speed it up just a little more.

  The process never ends. To accelerate it to the speed of light would take an infinite amount of energy, and is therefore impossible.

  In the same way, for a tachyon it takes more and more energy to slow it from above light speed. It would take an infinite amount of energy to slow it to the speed of light. Thus if both bradyons and tachyons exist, each is confined to its own velocity region. The speed of light is a "barrier" that forever separates the world of tachyons from the world of bradyons, and one can never become the other.

  This is all logically self-consistent, but a big question remains: How could one detect the presence of a tachyon? It used to be thought that any charged particle traveling faster than light would emit a particular radiation, known as Cherenkov radiation; but this is no longer believed to be the case. The simplest way to detect a tachyon's presence is through a time-of-flight test: if two particle detectors each register an event, and the distance between them is so large and the times so close that only a particle travelling faster than light could cause them both, then we have a candidate tachyon.

  This is a suspect mechanism for detection. If two people in a household come down with influenza within an hour of each other, we do not conclude that the incubation time for influenza must be one hour or less. Almost certainly, they both caught the flu from a third party. How would we ever know that a similar underlying cause did not lead to a false inference of tachyon presence?

  9.7 Wormholes and loopholes. Let us assume that tachyons exist. We then have a possible way of sending messages faster than light. That's useful, but it's not enough. We want to send people between the stars, fast enough to offer the writer some storytelling freedom. Tachyons won't do that, they are signals only. They also have an unfortunate property of allowing those signals to travel backward in time, which might alone be enough reason to avoid them.

  Where do we turn for plausible physics, a loophole that will allow us go faster than light without using tachyons or the still-unexplored world of quantum teleportation?

  As is so often the case, we turn to the work of Albert Einstein. Thanks to the general theory of relativity, the structure of the universe is not a single, simply-connected region of space and time. As we saw in Chapter 3, no information from inside a black hole can ever reach the rest of the universe. This is still true, even when we allow for the Hawking evaporation process. Thus a black hole provides, in a very real sense, an edge of the universe. If black holes are common, then the whole universe has a curious kind of Swiss-cheese structure of holes and real (i.e. accessible) space.

  Furthermore, there are regions close to a rotating black hole where very strange things can happen, at least in theory. Spacetime near the ring singularity of a kernel seems to be multiply connected. In other words, if you go close enough to the singularity, you may suddenly find yourself elsewhere, having been transported through a kind of spacetime tunnel. One problem is that you are likely to appear not only elsewhere, but elsewhen. The transport mechanism may also serve as a time machine.

  Other forms of spacetime tunnels, known popularly as "wormholes," have been developed by Kip Thorne and fellow-workers at CalTech. We say "developed," but of course the development so far is purely conceptual. It calls for extracting a minute black hole from the enormous numbers continuously appearing and disappearing at distances of the Planck length (again see Chapter 2). The trick is then to stabilize one—it will try desperately to disappear—and inflate it to a size useful for transmission of human-sized objects. This calls for materials far stronger than anything we have today, although our positronium-positronium bonds of Chapter 5 are taking us in the right direction. The magnified, stabilized wormhole can then serve, like our kernel ring singularity, as a way to travel between distant points without traversing intermediate "normal" space. Again, there seems to be a substantial danger that the traveler will appear in elsewhen.

  TABLE 9.1

  Frank Drake's proposed message to the stars.

  11110000101001000011001000000010000010100

  10000011001011001111000001100001101000000

  00100000100001000010001010100001000000000

  00000000001000100000000001011000000000000

  00000001000111011010110101000000000000000

  00001001000011101010101000000000101010101

  00000000011101010101110101100000001000000

  00000000000100000000000001000100111111000

  00111010000010110000011100000001000000000

  10000000010000000111110000001011000101110

  10000000110010111110101111100010011111001

  00000000000111110000001011000111111100000

  10000011000001
100001000011000000011000101

  001000111100101111

  Got it? No, neither could I.

  CHAPTER 10

  Deus Ex Machina:

  Computers, Robots, Nanotechnology, Artificial Life,

  and Assisted Thought

  10.1 Computer limits. This chapter of the book is a hard one to write. The reason is simple: if we are given a large set of data points on a graph and asked to say where the next points are likely to lie, we can do a fairly good job. We fit curves to the given points, and extrapolate. If we have just a couple of points, however, the task is practically impossible. We lack the information to decide which fitted curve is appropriate, and the next data points could lie almost anywhere. This is true in the best of circumstances, where no unexpected development puts a singularity on the time line and makes extrapolation impossible.

  Astronomy is as old as human history, and probably older. Mathematics, physics, and chemistry go back at least to Archimedes, in the third century B.C. Biology and medicine certainly predate Hippocrates, who lived around 400 B.C.

  Computers, even if we are generous with our chronology, stretch back at most to about 1832, when Charles Babbage began to formulate the ideas for a computing machine—an "analytical engine"—which he was never able to build. In terms of practical experience as to what a general purpose computer can do, we are limited to the half-century since ENIAC began operations in 1946.

  With a short history, our curve of projection can go almost anywhere. But computers, and what they may lead to, form so important a part of the human future that they cannot be ignored. We must assess where we are today, and where computers can plausibly go in the next decades and centuries. To see how computers have changed the science fiction world, recall that Heinlein and Clement, to name just two of the field's most famous and scientifically responsible writers, assumed that their heroes would be carrying slide rules around with them in the far future. The slide rule went the way of the dodo in the early 1970s.