http://youtu.be/AHr-iC93PPI My friend Chris Thompson just posted this video in which his 3D printer is printing a replacement part for itself… while the very part which he is trying to replace is about to fail.

It really points to some of the more interesting implications of a do-it-yourself 3D printer. From a price point and ability-to-spread perspective, a DIY 3D printer really has to be focused on printing as much of itself as possible. This is because self-printing:

• Brings the cost of parts down significantly; it’s not clear what the alternative could be (though laser-cut parts seem to be the main mass-produced alternative at this point), but an alternative would likely require investment in relatively expensive equipment (again e.g. a laser cutter)

• Stimulates interest - if you could only use a 3D printer for other objects, my guess is that interest might decline significantly since a large number of RepRap users spend a lot of time printing new parts

• Spreads out R&D costs and enables more rapid, smaller, testable improvements. To be clear, in the video Chris is printing out a critical (perhaps the critical) part of a 3D printer, which is the extruder that lays down the layers of hot plastic.  There are dozens of versions of the extruder available, and people seem to move slowly and gradually toward ones that are improvements on previous designs - more torque, more reliability, simpler, etc. The ability to self-print these parts allows for a lot more experimentation and testing, which would otherwise be very difficult to do and to spread. This is true for lots of other parts as well, including the software. And in fact, since the whole printer is made with commodity parts, a lot of experimentation can happen elsewhere also.

But I’m trying to think of any other systems that allow for actors within the system to manufacture their own upgrades. I can think of 2:

• Living things, but microscopic organisms (such as asexually- reproducing bacteria) in particular. Where this comparison fails is that an individual cell usually does not get upgraded - instead its descendants have mutations that may or may not be improvements. Still, evolution and variation result in lots of internally-manufactured changes that are tested, and can spread if successful.

• Certain social organizations. I’m not sure about this one, but look at democracy, for example. A democratic system can, wholly within the context of itself, test out and manufacture upgrades such as changes to procedures, rules, and laws. A market is similar - the stock market, for example, can change the rules to disallow short selling or certain types of trades, and these rules can spread to other stock markets if they help realize efficiency (or other criteria).

I don’t include humans themselves in this list because, at least as of now, we can’t manufacture our own upgrades (we have no power by ourselves to make changes to our physical functionality, though we can recruit others to do so).