3.1 Space Projects
1. Space Construction
Building stuff in space is one of the uses for this type of machine intelligence that I find most exciting. I am a strong believer in the maxim "never put all of your eggs in one basket." Humans need to colonize space and other worlds to help insure our survival and growth. And of course, it is just really cool and exciting. However, it is currently just way to expensive to do anything in space. It is so expensive that only governments can really do anything there, and even then it usually takes groups of governments to get stuff done because of the horrendous expenses. A good example of this is the new space station Alpha. Until space can be commercialized so that companies can make a profit, and average citizens can come and go freely and cheaply, then space will remain just out of reach. This page talks about two of the multitude of possible uses for machine intelligence in space. Of course, any large scale operations in space will depend critically on new technologies that will allow easy and inexpensive access. So until we get some space elevators or more efficient boost systems, allot of this stuff will stay in the realm of the dream world.
2. Spinning Out A Space Station
Constructing things in space is a very difficult proposition. Building the international space station has been an enormous undertaking. All of the major pieces of the station were constructed as modules on earth, flown up separately, and then hooked in place. And during that time, what is quite possibly the most expensive construction crew that ever existed, the astronauts connected each piece as it was placed in orbit. Only so many of these highly trained people were available to work at any given moment, and then only for very short periods of time. An easier and far more efficient means of construction is needed. And intelligent machines are the answer. The machines would be able to work full time, 24 hours a day, 7 days a week, on any projects, with minimal guidance from humans. Furthermore, they also hold the potential to solve another problem. The fact that each piece had to be manufactured on earth and sent up whole. If raw materials could be sent up aboard numerous smaller automated rockets, then the whole endeavor would no longer be completely dependent on the space shuttle and everything would be far cheaper. However, that still leaves the problem of how do you use the raw materials to build a space station? The solution I propose is to borrow an idea from spiders. Imagine a colony of spider robots that could spin out a space station just like real spiders can spin out a web to catch flies. Each spider would have a tiny factory in it that could produce a composite polymer of some kind. This could be done through a chemical reaction, or through physical nanomachines. It would then wind a bunch of these fibers over and over to form an incredibly strong composite wire that it could spin out just like real insects can spin out silk. The colony of spiders would then work together to build a complete and whole structure like a cocoon. Each laying down wires over and over again on a simple framing structure so that you end up with something that resembles the way cellulose makes up the body of plants. Other robots would have the ability to secrete sealants and plastics and such that would be able to complete the structure and make it space worthy for humans. But all that would need to be sent up are the robots and materials. Eventually it would even be possible for the robots to spin out nanotubes instead of composites. This would create incredibly strong structures that would no longer be so limited in size. Each piece of the international space station is limited by the cargo space available to lift it up in the space shuttle. But with this system you could build space stations of any size or shape as long as you could get enough robots and raw materials up there. Of course this technique could also be used to build spacecraft as well. And the craft it produced would be far superior to any jerry rigged contraption made up of separate pieces because it would be all one piece that is far stronger.
3. Colonizing Mars and the Moon
Insects are capable of amazing feats of construction. From things like the beehive and the ant hill to the termite mounds. They can build very large and complex structures even though each insect is individually not that smart. Ants for instance are almost completely hard wired. The behaviors they display in a given situation are determined largely by their genes and their birth. So how can such simple animals produce such stunning constructions. The answer is once again emergent behavior. It is the dynamic interaction of large groups of individuals that can perform such tasks. The same principle can be used with robots. Imagine that several landings have occurred on mars with raw materials, machines and finally a colony of robots with a number of different types of robots. One or more of the landing crafts would have the power systems. This would probably be a small nuclear reactor that the robots would treat as food. When the batteries on a robot are getting low it would become hungry and actively search for a food source to eat. It would find the reactor and hook into a receptacle to recharge itself. This is exactly like what the virtual insects do in the simulator presented here. The reactor systems would have some kind of RF beacons that the insects could use to home in on, just like the insects here used chemical odors. These insect robots would not be guided by humans. Humans would have no interactions with the robots unless there was a problem of some kind. Each individual robot would be responsible for navigating the landscape, finding a food source and eating. Once it has a full supply of energy it would once again go back to its normal job function. Upon emergence there will be a few different systems that will be needed immediately. So some of the insects will need to go and find the machines for these systems in other landers, carry them back to the construction site and set them up. Any constructions on mars will almost certainly be underground because of the violent dust storms and intense radiation due to the lack of an ozone layer. So that means digging. Since they will be doing all of that digging, the dirt might as well be used for something constructive. Why not make adobe bricks from them. One of the machines that will need to be setup will be the one that makes bricks. As each of the dirt hauler insects gets a full load it will dump it into the input of this machine. It will use some water and some power from the reactor to make a brick. Another couple of insects will be in charge of taking the newly formed bricks and stacking them for later use. While some of the insects are setting up machines, another group will begin breaking ground for the new habitat. There will be several different types of insects specially built to do digging and hauling of dirt. And just like bees can build incredibly complex and symmetrical hives, these insects could build a big hole in the ground with access ramps and the necessary scaffolding. There might be special insects that could bore small holes and then emplace small explosive rounds that would detonate to produce new loose dirt. Then others could come in and haul away the free dirt to be turned into bricks. Once they had built a hole that is big enough, the diggers would leave to begin a new hole somewhere else and the builders would take over. Each builder would retrieve a brick and emplace it to form walls in the new hole. They would be able to excrete sealants and stuff to act as mortar and to make it air tight, and would use the scaffolding and access ramps left by the diggers to build the whole thing room by room and floor by floor. Finally, not all of the dirt that was dug up will be used to make bricks. Some of it will be saved and used to fill back in around the structure so that it is once again buried underground. Next, the insects would retrieve other machines and parts from the landers and move them into newly constructed habitat. Things like air locks, computers, a power core, doors, furniture and so on. Also, hopefully we will discover that there is ice or water in places on mars. There is some good evidence for this, but no hard proof as yet. If so, then one of the first machines that would be brought online is one that takes in that water and splits it to form oxygen and hydrogen. The hydrogen will be stored for fuel, and the oxygen will be stored and used to generate other important gases like carbon dioxide so the new habitat will have an atmosphere once it is completed and airtight. Since there will be no organic organisms in the habitat then the atmosphere should stay pretty much as it is at first. Once the atmosphere is in place then the insects will construct the hydroponics bay. They will setup all of the equipment needed to begin growing a variety of new plants. Some type of system will be needed to maintain the CO2 levels for the plants, and a new type of gardener robot would be needed to tend to the plants and make sure they stay alive and grow. At that point you have a complete habitat ready for human arrival. They will have fuel, power, food, oxygen, water, and a place to live. We could just land and move in. Meanwhile, the little robots move to the next location and start all over again with a fresh set of raw materials that have just landed.
4. Problems To Overcome
Everything in the above description is possible. Not in a hundred years, or fifty years, but in ten to fifteen years at maximum. The simple simulated insect discussed on this website is already able to do many of these things. But what other things need to be accomplished to make this dream a reality? Well, the first steps will be those that I have outlined for my short term plans. You must first make this current simple system more realistic with a physics based simulation. Then build a robot and use the control system in it to get similar simple behavior. Next, use evolution to grow brains for the insects to get more complex and dynamic behavior. And then begin working on simulating groups of these insects to try and evolve insect brains that cooperate like bees and ants do. Also, the insects will need vision systems of some kind. Using antennas alone will simply not cut it. However, that does not mean that they will need vision systems as complicated as a humans. A bee has a visual system that is nowhere near as complicated as higher animals. Yet it is able to successfully fly through the environment without running into things. It can see predators and avoid them, and it uses its vision to home in on flowers where it can get its nectar. Everything here is doable in the near term.