Science-fiction stories about starships usually depict the universe as a collection of stars and planetary systems separated by vast stretches of empty space. The space between stars is imagined to be filled with dilute interstellar gas and nothing else. The real universe is much more interesting. The real universe contains a multitude of objects of various sizes, giving interstellar travelers places to stop and visit friends and collect fresh supplies between the stars. We know almost nothing about these objects except for the fact that they exist. We know that in the space around our own planetary system there are two populations of comets, known as the Kuiper Belt and the Oort Cloud.

The Kuiper Belt is the source of short-period comets, and the Oort Cloud is the source of long-period comets. We know that they exist because the comets which we see coming close to the sun are visibly disintegrating and cannot survive for a long time. The tail which makes a comet beautiful is proof of its mortality. Meteor showers are the debris marking the graves of dying comets. To keep new comets appearing at the observed rate, the source populations must be large, of the order of billions of comets of each kind. A few of the biggest and closest objects in the Kuiper Belt population can be directly observed, orbiting the sun with orbits concentrated around the plane of the planets. The brightest and most famous of these objects is Pluto. The Oort Cloud is invisible from the earth. It is a spherical population of objects at much greater distances from the sun, loosely attached to the sun by weak gravitational forces.

There is no reason to believe that the space between the Oort Cloud and the nearest stars is empty. We know that a large fraction of all stars are born with planetary systems. It is also likely that large numbers of planets are born unattached to stars. Furthermore, we know that the normal processes of formation and evolution of planetary systems result in ejection of planets and comets from the systems. As a result of these processes, the universe probably contains more unattached planets than stars, and billions of times more unattached comets. The space between our solar system and the nearest stars is probably infested with unattached planets and far more numerous unattached comets. In addition, there may be other objects of intermediate kinds which we have not yet observed, from snow-balls to black-dwarf stars. It is conceivable that some of the intermediate objects might be alive, a population of mythological monsters making their home in space.

The existence of abundant way-stations between the stars is likely to have a decisive influence on the development of starships. We shall not jump in one huge step from planetary to interstellar voyages. We shall be exploring one group of objects after another, first the Kuiper Belt, then the Oort Cloud, then a string of further-out oases in the desert of space, before we finally come to Proxima Centauri. In the history of mankind on this planet, there were two very different kinds of explorers who learned to navigate the oceans. There were the European navigators who sailed from fixed bases in Europe to destinations in America and Asia and came home to Europe with loot from their trade and conquest. Columbus was typical of these explorers, making three voyages back and forth across the Atlantic. But before the Europeans, there were Polynesian navigators who built canoes to sail long distances on the Pacific and populated the Pacific islands from Asia to Hawaii and New Zealand. The Polynesians did not have home bases in Asia and America, and they were not interested in sailing all the way across the ocean. They made their voyages from island to island, stopping to make a new home when they found a new island suitable for raising their crops and pigs and children.

The Polynesians were navigating the Pacific for a thousand years before the Europeans crossed the Atlantic. Island-hopping came first, intercontinental voyages later. It is likely that the future of our traveling beyond the Solar System will follow the same pattern. The evolution of starships, like the evolution of Polynesian canoes and European galleons, will proceed by a process of trial and error. Unattached comets and planets will be like the islands in the Pacific Ocean. We will begin like the Polynesian navigators, modestly. Developing starships one step at a time, we can learn by trial and error how to do the job right. Perhaps, after a thousand years, we will be ready to build grand super-highways conveying traffic along non-stop routes from star to star.

Two things are needed to make starships fly, a place to go and a way to get there. The first problem is mainly a problem of biology, the second a problem of engineering. Let us look at biology first. To have a place to go, we must learn how to grow complete eco-systems at remote places in the universe. It is not enough to have hotels for humans. We must establish permanent ecological communities including microbes and plants and animals, all adapted to survive in the local environment. The populations of the various species must be balanced so as to take care of each others’ needs as well as ours. Permanent human settlement away from the earth only makes sense if it is part of a bigger enterprise, the permanent expansion of life as a whole. The best way to build human habitats is to prepare the ground by building robust local ecologies. After life has established itself with grass and trees, herbivores and carnivores, bacteria and viruses, humans can arrive and build homes in a friendly environment. There is no future for humans tramping around in clumsy space-suits on lifeless landscapes of dust and ice.

The recent revolution in molecular biology has given us new tools for seeding the universe with life. We have learned to read and write the language of the genome, to sequence the DNA that tells a microscopic egg how to grow into a chicken or a human, to synthesize the DNA that tells a bacterium how to stay alive. We have sequenced the genomes of several thousand species. The speed of sequencing and of synthesis of genomes is increasing rapidly, and the costs are decreasing equally rapidly. If the increase in speed and the decrease in costs continue, it will take only about twenty years for us to sequence genomes of all the species that exist on our planet. The genetic information describing the entire biosphere of the planet will be available for our use. The total quantity of this information is remarkably small. Measured in the units that are customary in computer engineering, the information content of the biosphere genome amounts to about one petabyte, or ten to the power sixteen bits. This is a far smaller amount of information than the data-bases used by enterprises such as Google. The biosphere genome could be embodied in about a microgram of DNA, or in a small room full of computer memory-disks.

Looking ahead fifty or a hundred years, we shall be learning how to use genetic information creatively. We shall then be in a position to design biosphere populations adapted to survive and prosper in various environments on various planets, satellites, asteroids and comets. For each location we could design a biosphere genome, and for each biosphere genome we could design an egg out of which an entire biosphere could grow. The egg might weigh a few kilograms and look from the outside like an ostrich egg. It would be a miniature Noah’s ark, containing thousands or millions of microscopic eggs programmed to grow into the various species of a biosphere. It would also contain nutrients and life-support to enable the growth of the biosphere to get started. The first species to emerge from a Noah’s ark egg would be warm-blooded plants designed to collect energy from sunlight and keep themselves warm in a cold environment. Warm-blooded plants would then provide warmth and shelter for other creatures to enjoy. In this way, life could be seeded in great abundance and variety in all kinds of places, traveling on small space-craft carrying payloads of a few kilograms. Since life is inherently an unpredictable phenomenon, many of the biospheres would fail and die. Those that survived would evolve in unpredictable ways. Their evolution would continue for ever, with or without human intervention. We would be the midwives, bringing life to birth all over the universe, as far as our Noah’s ark eggs could travel.

The second problem, the problem of engineering, is to build machines that can take us from here to there. To have space travel over long distances at reasonable prices, we must build a public highway system so that the costs of the initial investment can be shared by a multitude of users. A public highway system in space will require terminals using sunlight or starlight to generate high-energy beams along which space-craft can fly. The beams may be laser-beams or microwave-beams or pellet streams. The massive energy-generating machinery at the terminals remains fixed. The space-craft are small and light, and pick up energy from the beams as they fly along. Unlike chemical or nuclear rockets, they do not carry their own fuel. For the system to operate efficiently, the volume of traffic must be big enough to use up the energy of the beams. Space-craft must be flying along the beams almost all the time. As with all public highway systems, the system can only grow as fast as the volume of traffic. The cost of travel will be high at the beginning and will become low when every terminal is crowded with passengers waiting for a launch.

In every public transport system, things work better if we build separate vehicles for passengers and freight. On the roads, cars for passengers and trucks for freight. On the railroads, fast short trains for passengers and slow long trains for freight. The Space Shuttle was a system designed to put passengers and freight on the same vehicle, and that was one of the reasons why it failed. It was supposed to be cheap and safe and reliable, with frequent flights and a high volume of traffic, and it turned out to be expensive and unsafe and unreliable. The public highways of the future will be like roads and railroads and not like the Shuttle. But the relation between passengers and freight in the future will be the opposite of what it was in the past. In the past, humans were small and light, freight was big and heavy. Cars were small and agile, trucks were big and clumsy. In space today, this relation between human passengers and freight is already inverted. Because of the miniaturation of instruments and communication systems, unmanned spacecraft have become smaller and lighter than manned spacecraft. Payloads of unmanned missions have remained roughly constant while their performance and capability have improved by leaps and bounds. Payloads of manned missions have remained larger while politicians fail to decide what they are supposed to do.

In the future, when missions go beyond the solar system, the difference between passengers and freight will become greater. Freight will no longer be bulk materials such as fuel and water. Freight will be information, embodied in ultralight computer memory or in DNA. Freight will be several orders of magnitude lighter than human passengers. Payloads of unmanned missions may be measured in grams, while payloads of manned missions will always be measured in tons. As a result, the public highway system will consist of two parts, a heavy-duty system transporting human passengers between a small number of metropolitan human habitats, and a light-freight system transporting packages of information along a wider network of routes to more distant destinations. A typical light-freight mission might be like the Starwisp proposed by Bob Forward. The Starwisp is an ultralight sail made of fine wire mesh, driven through space by a high-power beam of microwaves. The wire mesh is not only the vehicle but also the payload, carrying sensors to explore the environment and transmitters to send information collected by the sensors to humans far away. Starwisp could also be a vehicle for carrying Noah’s ark eggs to bring life to remote places. It is likely that the travel-times of voyages will become longer than a human life-time. After life has spread that far, it will no longer make sense for humans to travel with it. Instead of imprisoning human travelers for a lifetime in a space-craft, it would make more sense to load the space-craft with a few human eggs, which could grow into humans at the destination. In the end, we would populate the galaxy by broadcasting the information required for growing humans, rather than by carrying deep-frozen human bodies for thousands of years.

When we are thinking about the spread of life into the universe, the most important fact to remember is that almost all the real-estate in the universe is on small objects. Real-estate means surface area. The universe contains objects of all sizes. Most of the mass and volume belong to big objects such as stars and planets. Most of the area belongs to small objects such as asteroids and comets. Most of the life will have to find its home on small objects. The majority of small objects have three qualities which make them unfriendly to life. They are far from the sun or other stars, they have no atmosphere, and they are cold. In spite of those disadvantages, they can be seeded with life. They can support biospheres as diverse and as beautiful as ours.

The key technology for bringing life to small cold objects in space is the cultivation of warm-blooded plants. Warm-blooded plants are more essential to the ecology of cold places than warm-blooded animals are to the ecology of our warm planet. Life on earth might have evolved happily without birds and mammals, but life in a cold place could never get started without warm-blooded plants. Two external structures make warm-blooded plants possible, a greenhouse and a mirror. The greenhouse is an insulating shell protecting the warm interior from the cold outside, with a semi-transparent window allowing sunlight or starlight to come in but preventing heat radiation from going out. The mirror is an optical reflector or system of reflectors in the cold region outside the greenhouse, concentrating sunlight or starlight from a wide area onto the window. Inside the greenhouse are the normal structures of a terrestrial plant, leaves using the energy of incoming light for photosynthesis, and roots reaching down into the icy ground to find nutrient minerals. Since there is no atmosphere to supply the plant with carbon dioxide, the roots must find mineral sources of carbon and oxygen to stay alive. We see in the light emitted from comets, as they come close to the sun, that these icy objects contain plenty of carbon and oxygen as well as nitrogen and other elements essential to life.

The embryonic warm-blooded plant must grow the greenhouse and the mirror around itself while still protected within the greenhouse of its parent. The seeds must develop into viable plants before they are dispersed into the cold environment. These plants must be viviparous as well as warm-blooded. It seems to be only an accident of evolution on our own planet that animals learned to be viviparous and warm-blooded while plants did not.

The optical concentration that the mirror must provide will depend on the distance of the plant from the sun or star providing the energy. Roughly speaking, the optical concentration must increase with the square of the distance from the source. For example, if the plant is on the surface of Enceladus, a satellite of Saturn at ten times the Earth’s distance from the sun, the intensity of sunlight is one hundredth of the intensity on Earth, and the optical concentration must be by a factor of a hundred. If the plant is in the Kuiper Belt at a hundred times the Earth’s distance, sunlight is reduced by a factor of ten thousand and the mirror must concentrate by a factor of ten thousand. Existing biological structures can do much better than that. The human eye is not an extreme example of optical precision, but it can concentrate incoming light onto a spot on the retina by a factor larger than a million. That is why staring at the sun is bad for the health of the eye. A mirror as precise as a human eye would be good enough to keep a plant warm at a distance ten times further from the Sun than the Kuiper Belt. Eagles and hawks have better eyes than we do, and a simple amateur telescope costing less than a hundred dollars is better still. There is no law of physics that would prevent a warm-blooded plant from growing a mirror to concentrate enough starlight to survive anywhere in our galaxy. The main difficulty in achieving a high concentration of starlight is that the mirror must track the source accurately as the object carrying the plant rotates. The plant must be like a sun-flower, tracking the sun as it moves across the sky. If high accuracy is needed, the plant must grow an eye to see where it is pointing.

These speculations about viviparous plants and Noah’s ark eggs and life spreading through the galaxy are my personal fantasies. They are only one possible way for the future to go. The real future is unpredictable. It will be rich in surprises that we have not imagined. All that we can say with some confidence is that biotechnology will dominate the future. The awesome power of nature, to evolve unlimited diversity of ways of living, will be in our hands. It is for us to choose how to use this power, for good or for evil.