technology

self-expanding domains

I am talking about a domain, or category of things that are made with certain other things, perhaps a particular set of tools. They take an input, and produce an output within the domain. An example would be woodworking–things people use woodworking tools to make. You take wood from trees, and use woodworking tools to make wooden toys, furniture, or signs.

What makes a domain self expanding is if the tools/implements/whatever are in the domain itself. You can use woodworking tools to make the handle of a saw, but not a functional saw. So woodworking is a partially–and weakly–self-expanding domain.

There’s only one completely independent self expanding domain I could name, and that is life. Living things can metabolize non-living matter and produce living matter. Mushrooms and plants can use chemical and solar energy to break up rocks and turn them into more mushrooms and plants, and more bioavailable dirt. Also, by definition, any living thing has the capacity to reproduce itself, on top of whatever else it produces. Species within life count as self-expanding domains, as does life overall.

When the inputs are within the domain we call it domain cannibalism, and when they come from outside the domain we call it genuine expansion. For example, when you eat plants, for the overall domain of life, that would be domain cannibalism, but for the subdomain of human beings, it would not. Elements brought into the domain we call vitamins.

Some domains are indirectly self-expanding. Viruses do not reproduce directly, but they hijack living cells to produce more viruses.

Other domains are only partially self-expanding. Other than life, all the ones I can think of require human operation to force the action forward. They need to produce something that humans want other than the tools, so humans will continue to propagate those tools and and maintain and expand the domain’s capacity for expansion. And, of course, they need to reproduce, either of their own volition or with human operation.

The reproducing elements we call domain seeds. The products that do not reproduce we call domain flowers, in analogy to the way flowers and bees interact to reproduce more flowers. Flowers attract bees; useful products/domain flowers attract humans.

Early humans saw life propagating itself, and noticed that seeds from a plant grew into the same plant. So they planted some of the ones they liked to eat, chose the best ones to replant, and the first human-operated self-expanding domain was born: agriculture. Some of our plants would continue replicating themselves without us, but many wouldn’t. It takes a lot of human work to keep that self-expanding domain on track, producing and reproducing.

Machine and metalworking tools are almost entirely self-expanding, though they need a human operator. I’ve heard both the lathe and the mill  referred to as the ‘mother tool:’ they can be used to make the entire shop of tools, including themselves. That may not be the most efficient way for them to reproduce, but even the most efficient way involves their use: they are needed to produce the molds and stamps that produce their pieces. They form the basis of industrial production.

Software is built using other software (compilers and various packaging tools), but all software depends on computer hardware to run on, and humans to use them. That hardware is produced in a process that involves software, computer hardware, and a range of other tools.

Computer hardware is also arguably partially self-expanding at this point. Early semiconductors could be designed using pen and paper, but now with billions of circuit elements on a given chip, that’s just unmanageable. Also, mass production of semiconductors requires fairly sophisticated automated processes.

So there are two self-expanding domains that intersect, computer hardware and software. Software helps hardware create more hardware, and hardware gives software a place to run and do all the things it does, including make more software. Together they form the basis of an information based economy.

These domains also support evolution: as domain seeds are used to produce more domain seeds, they can be used to improve and test designs, advance the state of the art, and accelerate change as seed designs are refined. Repraps are an attempt to make that loop more explicit, and to use the open-source model to broaden involvement in that process and accelerate it.

Also, note: agriculture, industrialism, information technology: we are distinguishing the core of epochal technological transformations. When a new self-expanding domain is invented, it’s transformational to the economy.

This a materialist idea: it’s about things making other things, or people making things with other things. You could think of it as a material extension of memetics. I’ve tried to extend it to economics or culture, but I think memetics already explains that.