This is my second post filling in the details of a Star Trek vision. In Star Trek  replicators make everything from parts for spaceships to food for human consumption. One would assume that a replicator can also make the parts necessary for more replicators, thus replicating itself. As such it is an artistic vision of what von Neumann famously conceived as the universal constructor, a machine that can replicate itself and through the process of mutation grow to arbitrary complexity. While the exact functioning of the Star Trek replicator is never spelled out it seems to rely on accessing energy too cheap to meter. So suppose that we made a lot of progress on energy, how could we pursue a vision of fully automated production.

While we are a long way away from a compact replicator, we have made tremendous progress in manufacturing. Because only a small percentage of the population directly work in manufacturing, much of this progress has been invisible to most of the public. Today we have highly capable 3D printers that can additively make production parts in a wide range of materials including metals and ceramics. We have highly automated CNC machining cells for more traditional techniques (which remove material). We have the ability to automate metal folding and plastic injection molding. We can print PCBs and place parts on them in a highly automated fashion. And we now have recently unlocked much better control of robotic arms with rapid progress on hands also thus allowing us to automate assembly of all of these parts into a finished product. And we have robots that can supply machines with materials, move parts between machines and assemble parts into products.

Looking around a house, an office, or even a factory, it is hard to come up with objects for which we cannot envision fully automated production with today's technological capabilities. We are not talking here about stuff that's hidden deep in a research lab somewhere but rather these are capabilities that are in production somewhere already or about to go into production soon. Or in the inimitable words of William Gibson "The future is already here. It is just not yet evenly distributed."

We should be pushing these capabilities forward as rapidly as we can. Why? Because it allows us to make technology broadly deflationary without having to rely on cheap, possibly slave, labor in far away countries. Because many of these techniques bring with them increased flexibility they also tend to reduce the scale necessary for cost effectiveness. The limit of course once again being the replicator (which can make anything at the marginal cost of energy, which is assumed to be near zero). Increased flexibility and smaller scale are desirable because they are at the heart of innovation.

So what do we need to go harder on this? At first blush it would seem that we don't need much from the government here, but there is a lot that would help to accelerate and secure highly automated manufacturing.

The first point is of course that we need to dramatically drive down the cost of energy.

The second point is that all of this technology depends on chips (micro controllers are everywhere) and we don't (yet) have nearly enough fab capacity domestically to make these. There is also still the dependence of the fabs on ASML and on Zeiss. I am less worried about the geopolitical aspects here but rather these are centralizing forces that slow down innovation. Imagine if we had microfabs everywhere that could tape out custom chips overnight. While there are some startups in this area, research into decentralizing chip production should be much more heavily funded by government.

The third point is one that I make in The World After Capital. Labor needs to be expensive so that automation truly pays off. And then we also need money in the hands of people so that they can buy these products. Universal Basic Income (UBI) accomplishes both of these at the same time, which is why it has to be a central policy going forward. To be clear, UBI is not a panacea and making it effective requires far reaching other changes from the tax code to the education system. Also for some adjustment time period tariffs might need to play a role here to adjust for subsidized labor and energy, as well as for missing carbon taxes.