I was watching ‘Megastructure Death’ this afternoon on SFIA and got to thinking about megastructures again, and clanking replicators from a couple days ago. This is technology we can expect to see deployed to the Moon in the very near future, although we need to finish designing it! 8-P By which I mean a discrete industrial base which can make all the parts to make all the parts to make itself, probably leaning very heavily on additive manufacturing, but traditional processes can get the job done, because they already have. Like with molecular nanotechnology where life itself can be considered a proof of concept, the entire GWP can be considered a clanking replicator, only one spread out across the entire globe producing some fraction of roughly ~$100 trillion worth of goods and services. We can grant that something a bit smaller will do the job, on the scale of a regional economy, like a city on Mars. If a medium-sized country with a trillion dollar industrial sector and 100 million people, for one digit accuracy, with a hundred tonnes worth of physical capital for 10 billion tonnes of industrial base, we can consider that the minimum which could replicate itself, in the aftermath of a global disaster which leaves that country (mostly) intact, for instance...

To get to the point and make some wild-ass guesses, we could maybe distill this down to (10^10)^½ or 100,00 tonnes for a research cost of (10^9)^½ or $31,000, not very much as I imagine we would not need to come up with anything new to get such a wonder factory. We want something more compact, say 10^10/3 or 10^3.33 tonnes, a couple kilotonnes at the research cost of 10^9*¾ or 10^6.75, a bit under $5 million. Even if this is more like $5 billion, that’s still less money than $20 billion to get the clanking moon replicator to the moon at $10 million/tonne, which should be able to manufacture an order of magnitude more stuff in the first year, $200 billion worth of goods at $10 million/tonne, so we’re getting to the point where the potential trillion dollar orbital manufacturing capacity of the middle 21st will make that profitable and therefore possible.

Once this exists and gets better with every passing year we would see it deployed to the earth over time. A two kilotonne sea borne clanking replicator, say, or one set up at Yellowstone to build and be powered by a geothermal tap for taming and turning the supervolcano into an asset instead of an existential threat. A 2 kilotonne clanking replicator would also be something we could live with or in, at that 100 tonne/person quote; what with 10% standard depreciation and 24 tonnes per tonne per year, a clanking replicator could maintain a half megatonne of arcology for 5,000 people at 100 tonnes per person. This is less than half a percent of the mass of the arcology, so the clanking replicator could be considered to be built into the arcology a few times over for redundancy and not be painfully obvious.

At a $1 million/tonne, the CR would cost a couple billion, which only sounds like a lot of money until you consider that this is $2 million per household for that 5,000 person arcology. An awful damn lot of money for most, but not for the upper tier who are shelling out half a million and more for a new house, much more for a McMansion or condo. That and after a few years this tech would go for a few tens of millions. A billionaire could easily set up a self-sufficient compound anywhere in the world in the early days and eventually have a squatters’ village next door...