In 1969, after graduating from Harvard but before starting further study in astro-physics at Princeton University, I took a summer holiday in Europe and visited the Berlin Wall. It was the height of the Cold War, and the wall was then eight years old. Standing in it ominous shadow, I began to wonder how long it would last. Having no special knowledge of East-West relations, I hadn't much to go on. But I hit on a curious way to estimate the wall's likely lifetime knowing only its age.

I reasoned, first of all, that there was nothing special about my visit. That is, I didn't come to see the wall being erected or demolished--I just happened to have a holiday, and came to stand there at some random moment during the wall's existence. So, I thought, there was a 50 per cent chance that I was seeing the wall during the middle two quarters of its lifetime (see Diagram, below). If I was at the beginning of this interval, then one-quarter of the wall's life had passed and three-quarters remained. On the other hand, if I was at the end of of this interval, then three-quarters had passed and only one-quarter lay in the future. In this way I reckoned that there was a 50 per cent chance the wall would last from 1/3 to 3 times as long as it had already.

Before leaving the wall, I predicted to a friend, that it would with 50 per cent likelihood, last more than two and two-thirds years but less than 24. I then returned from holiday and went on to other things. But my prediction, and the peculiar line of reasoning that lay behind it, stayed with me. Twenty years later, in November 1989 the Berlin Wall cam down--unexpectedly, but in line with my prediction.

Intrigued that the approach seemed to work, I eventually set out its logic in Nature(vol 363, p315, 1993). There, instead of using the 50 percent mark, I adopted the more standards scientific criterion that the prediction should have at least a 95 per cent chance of being correct. This makes the numbers in the formula come out a bit different, but the argument remains the same. If there is nothing special about your observation of something, then there is a 95 per cent chance that you are seeing it during the middle 95 per cent of its observable lifetime, rather than during the first or last 2.5 per cent (see Diagram, p 38). At one extreme the future is only 1/39 as long as the past. At the other, it is 39 times as long. With 95 per cent certainty, this fixes the future longevity of whatever you observe as being between 1/39 and 39 times as long as its past.

36	New Scientist	15 November 1997

This formula can be used to amusing effect. Mathematicians Peter Landsberg, Jeff Dewynne and Colin Please of the University of Southhampton used it to predict how long the Britain's Conservative government would stay in power (Nature, vol. 365, p385, 1993). The Conservative Party had been in power for 14 years, and they estimated with 95 per cent confidence that it would be in power for at least 4.3 months but less than 546 more years. Sure enough the Conservative Party was ousted 3.5 years later.

Living Proof

This is all good fun. You can predict approximately how long something will last without knowing anything that its current age. But in the past few months, in the light of the spectacular success of NASA's Mars Pathfinder mission, I've been reminded of of a far more serious implication of this way of thinking. Applying it to the human race forces mt to conclude that our extinction as a species is a very real possibility, and that we had better take steps to improve our survival prospects before it's too late. Let me explain why I have such a sense of urgency, wand why we had better begin colonising space--and very soon.

In the 16th century, Nicolaus Copernicus pointed out that the Earth revolved about the Sun, rather than vice versa, and in one swift move, displaced humanity from its privileged place at the centre of the Universe. We now see the Earth as circling an unexceptional star among thousands of millions of others in our unexceptional Galaxy. This perspective is summed up more generally in the "Copernican Principle", which is the position that one's location is unlikely to be special.

Early this century, when astronomer Edwin Hubble observed approximately the same number of galaxies receding from Earth in all directions, it looked as if our Galaxy was at the exact centre of a great explosion. But reasoning with the Copernican principle, scientists concluded instead that the Universe must look that way to observers in every galaxy--it would be presumptuous to think that out galaxy is special. As a working hypothesis, the Copernican principle has been enormously successful because, out of all the places intelligent observers could be, there are only a few special places and many nonspecial places. A person is simply more likely to be in one of the many nonspecial places. but the Copernican principle doesn't apply only to placement of galaxies in space-- it works for placement of moments of time as well. Inset 1

What does it imply for Homo sapiens? We have been around for about 200 000 years. If there is nothing special about the present moment, then it is 95 per cent certain that the future duration of our species is between 1/39 and 39 times 200 000 years. That is, we should last for at least another 5100 years but less than 7.8 million years.

Since we have no actuarial data on other intelligent species, this Copernican estimate may be the best we can find. It gives our species a likely longevity of between 0.205 million and 8 million years, which is quite in line with those for other hominids and mammals. The Earth is littered with the bones of extinct species and it doesn't take much to see that we could meet the same fate. Our ancestor H. erectus last 1.6 million years, while H. neanderthalensis lasted 0.3 million years. The mean duration of mammal species is 2 million years, and even the great Tyrannosaurus rex lasted only 2.5 million years.

For us, the end might come from a drastic climate change, nuclear war, a wandering asteroid or comet, or some other catastrophe that catches us by surprise, such as a bad epidemic. If remain a one-planet species, we are exposed to the same risk as other species, and are likely to perish on the same timescale.

38	New Scientist	15 November 1997

Some people might think that the discoveries of our age--space travel, genetic engineering and electronic computers--place us in a special position. These breakthroughs, they might say, could lead us to spawn new intelligent species, including intelligent machine species, enhancing our chances of survival. But such thinking may raise false hopes. For, according to the Copernican principle, you are likely to be living in a century when the population is high because most people will be born during such periods. And since it is people who make discoveries, it is not surprising that you will live in a century when many interesting discoveries are being made. But your chance of being born 200 000 years after the beginning of your intelligent lineage, in the very century when a discovery is made that guarantees it a billion-year future, is very small, because a billion years of intelligent observers would be born after such a discovery, and you would be more likely to be one of them. If you believe that any current discovery will dramatically increase our longevity, you must ask yourself: why am I not already one of its products? Why am I not an intelligent machine or genetically engineered?

Act Now

In my 1993 Nature paper I estimated how long the human space programme, then 32 years old, would last. Since my paper was not likely to fall either in the first 2.5 percent or the last 2.5 per cent of the programme, I predicted with 95 per cent confidence that its future duration would be more than 10 months but less than 1248 years. The upper limit of 1248 years is the total number of future years of human space flight, regardless of how many periods of inactivity occur.

You might argue that here is no hurry to colonise space within the next century. Why not wait a few centuries until technology has become so advanced that colonising becomes easy? But if we lose the capacity for spaceflight before we've colonised--by the collapse of civilisation, loss of technology or diminished economic ability--then we've missed our chance. It's good that we went to the Moon in the 1960s. If we'd waited another 30 years hoping for an easier time of it, we might never have made it, as we now seem to have less money for such ventures. Inset 2

Unfortunately, I believe that we are likely to make precisely this kind of mistake. In 1969, Wernher von Braun had plans to send astronauts to send astronauts to Mars by 1982; in 1989 Geirge Bush proposed sending people there by 2019. This is an unfortunate trend, and I'm worried that the day may come when there is no one left alive who can say "I walked on the Moon". People who realise that colonising the Galaxy would be very beneficial to our survival have generally regarded such as inevitable. But it is not.

Since you are still on your home planet, the Copernican principle tells you that a significant fraction of all intelligent observers must also still be on their home planets (otherwise you would be special). This explains why we have not been colonised by extraterrestrials--a significant fraction of them are still sitting at home. I would be more confident about the future if we were members of a billion-year-old civilisation which had already colonised is galaxy. But our chance of colonising the entire Galaxy, increasing our current population by about a factor of a billion, is about one in a billion. Why? Because it would mean that you were born within the first billionth of all humans, which is exceedingly unlikely (my colleagues Brandon Carter, John Leslie and Holgar Nielson have reached similar conclusions). The fraction of all civiliastions that achieve galactic colonisation is likely to be small --otherwise, you would likely be living in such a civilisation now.

But that doesn't mean that we can't at least get off the planet and plant some colonies that will greatly enhance our survival chances. Colonisation, starting with Mars, should be our first goal. Space experiments could be geared towards making this possible. And if colonisation were the goal you would not have to bring the astronauts back from Mars--after all, that is where we want them. Instead, we could equip them to stay ("Escape from Mars", New Scientist, 28 June, p 24) and establish a colony at the outset--a good strategy if one is worried that funding for the space programme may not last.

So, we should be asking ourselves: what is the cheapest way to establish a permanent, self-sustaining colony on Mars? How many Saturn V-class launches would it take to land eight people o Mars with enough supplies to last them 30 years, with a reasonable recycling strategy? Could those people in 30 years build a habitat big enough for them and their 16 children using only indigenous materials? And in another 30 years, without further supplies from Earth, could those 16 children enlarge the habitat to include 32 grandchildren, and so on? The original mission could even include a large sample of frozen egg and sperm cells to supply additional genetic diversity.

What a bargain colonising space is--we send out a few astronauts, and their descendants do most of the work. Shouldn't we be trying, before it's too late?

15 November 1997	New Scientist	39

Diagaram 1

Diagaram 2

Inset 1

Inset 2