In any developing system, there needs to be goals, milestones, and measures of progress. In other chapters, we’ve discussed economic fitness models. These are much like the blunt instrument of natural selection, they lack the fine detail that an engineer would need. In this chapter, I examine degrees of autonomy as another measure of success.
The basic system of RASA would be considered the zero state. An empty box, able to move freely in all three axes of a pre-built framework, is the simplest form. Naturally, there are things that can be improved upon such a system, and any upgrade that increases the autonomy of such a system would be considered a measurable step towards a fully closed self-replicating machine. Also, it can be thought of as the removal of one task, however small, that human beings would have to perform for the system to operate. This is quite a broad definition, so I will try to provide examples and categories of degrees of autonomy.
Pure System Improvements:
First of all, there are improvements to the basic system that can increase autonomy, which can include:
Energy system improvements, such as charging stations throughout the grid. This would replace people manually providing batteries or fuel.
Control system improvements, where a person would not have to physically direct the small details of the system functions.
Fault detection systems, which would identify maintenance problems as they come up.
Payload to Basic System Processes:
Probably more importantly, is the category of degrees of autonomy that have to do with the ecology of the machinery that will ride within the RASA system (the payload), and how it relates to the RASA system itself. Each of these tasks could potentially be carried out by a custom machine nested within a cube. Some examples would include:
Construction machines that can assemble the grid framework by themselves, providing they are given a steady supply of the correct parts. Also, machines that can attach or assemble improvements to the basic grid, such as power lines and charging stations.
Construction machines that can build or assemble the empty mechanized boxes that drive around the grid.
Maintenance machines, which could deal with problems with the grid as well as mechanical problems and breakdowns of other machines. An ability for one machine to take another defective machine out of service and bring it to a service centre would be incredibly useful. I think even early systems should at least be designed with this ability in mind for the future.
Pure Payload to Payload Interaction Improvements:
Important degrees further down the road are improvements relating to how two or more payload machines can interact with each other, where the RASA system is just the passive carrier among them. Possible examples of this are:
Payload machines that can assemble and/or service other payload machines.
Payload machines that make parts for other payload machines. Depending on the construction of machines and parts, and then the parts for the machines that make the parts, and so forth down the line, this can easily develop into a complex economy of machines serving machines.
Machines that harvest raw materials and pass them on to other machines to make parts. Mining machines, gathering machines, and so forth. Machines that move and dispose of waste as well.
Cargo machines, which store, carry, and transfer different types of materials around for other machines to use.
Now that we have discussed some of the general types of improvements that would push the system towards a greater degree of self-replication, it’s possible to write down on paper and count how many degrees of autonomy a system has. I can envision a future, where someone is designing a RASA based industrial application, and uses degrees of automation to help pick and choose which machines would be needed to get the job done in the best and most economical way.
Using degrees of autonomy, it’s possible to measure the efficiency of a machine in degrees of autonomy per machine. For instance, replacing a machine that makes a special part with a machine that can make twenty different parts would be a great improvement. Degree density would be an obvious measure of efficiency, though not the only one. My guess is that measuring degrees of autonomy will see much more practical uses as the complexity of the system increases.
Unlike degrees on a compass, each degree of autonomy of a self-replicating system is unique and generally not replaceable by another degree, but this can change as things get complex. It’s easier to see this as an analogy to degrees of separation between people or other such separations on a network, in that each path between two points is unique. Just like interconnected groups of friends, or nodes on a network, there is often more than one path between points A and B. Once the RASA system becomes a sufficiently complex self-sustaining economy, there could easily be more than one way to perform any given task, and measuring how many machines or degrees of autonomy are needed for that task can help in deciding which is more efficient or economical.
In research and development, a future company can pick any specific degree of autonomy that isn’t built yet, and by developing a working system that satisfies that degree, they can turn a profit by supplying that to anyone who would need that for their own work. It would take a lot of guesswork out of venture capital because these kinds of degrees would resemble niches in biological ecosystem in that there’s a rule that something will eventually fill an empty niche. And in the bigger picture, a designer for a highly desired industrial process could break it down into various degrees of autonomy, and focus separate research teams to each that would come together in the whole without ever needing to be aware of one another, as long as a sufficient standard system is in place that allows them to inter-operate. I feel the ultimate flexibility and modularity that the RASA system provides will lead people in the future to wonder how they ever did without it in the same way that we now wonder how we did without the Internet.