Rise of the Replicators

We forget there was a time before all of us walked around with all of human knowledge in our hand.

But it was only yesterday.

The concatenation of the smartphone and the mobile Web and Google and Wikipedia allows us to inform every decision we make with the best available information. We can always leverage off the experience of others. Millions of others.

That we do not is a sign of inertia, that our behaviours have not yet caught up to our capacities.

That gap between what we can do and what we are actually doing, that’s something we don’t really talk about much.

Because while it’s obviously true for knowing — fake news is the shadow of a world where we can know anything — it is perhaps less obviously true for doing.

That we can do almost anything now.

It all circles back to the smartphone.

It’s been my great pleasure to be affiliated with the University of Sydney’s INCUBATE program over the last few years, mentoring startups.

Several of these startups came out of the Australian Centre for Field Robotics, run by Dr. Salah Sukkarieh. And at an INCUBATE event this winter, he showed the current work at the centre.

The Australian Centre for Field Robotics is doing some great work

It’s an agricultural robot designed to be both cheap and flexible enough for deployment in the developing world.

It’s basically two motorised wheels on an axle, with something that looks a lot like a selfie stick extending out from it.

That selfie stick holds a smartphone. That smartphone is doing all the work of controlling the robot, as well as providing the camera input the robot needs to examine crops for disease, pests, and growth.

The farming robot brings the connected intelligence of farmers and agronomists all over the world to the field, to the farmer — wherever there’s mobile broadband. In 2017, that’s almost everywhere in the world.

Best of all, that farming robot is simple enough that it can be sold for about two thousand dollars — and would likely be leased for use by a whole village of farmers who will all earn enough, using the robot, to easily make the lease payments.

All of this is possible because the brains of the operation — the smartphone — is manufactured by the billions every year, so the cost of connected computing and sensing has, over the last decade, dropped more than a hundred fold.

It’s that uptick in capacity — as exemplified by this farming robot — which I’d like to focus on.

This transition from a knowledge-rich to a capacity-rich culture has many different faces, but I want to focus on the capacities which most directly affect the lives of people: material capacities.

Let me start off with a thought experiment from the earliest days of computing.

John Von Neumann — the father of our modern computer architectures — described a ‘self-replicating machine’ — that is, a machine programmed to make copies of itself, which would then make further copies, and so forth.

That’s the way all living organisms work. It’s certainly more efficient to build a machine that can make copies of itself than to build all those copies.

Von Neumann’s work seemed theoretical — or the stuff of science fiction nightmares — but in 2005 the RepRap project kicked off at the University of Bath.

That project defines a 3D printer that can print pieces from plastic — including most of the components of a 3D printer.

It’s a printer that can — mostly — print copies of itself.

RepRap is a commons-based project.

As with Wikipedia, the participants pooled their knowledge, because they understand ‘a resource shared is a resource squared’.

Once a RepRap printer has printed most of the components for another RepRap printer, they need to be assembled — by a person — into a working RepRap printer.

That’s a delicate operation, and one that requires some experience. It’s also time-consuming.

What you’d really want is a RepRap that could print a copy, then assemble that copy.

There’s already a kind of machine that does it, it’s called a ‘pick and place’ robot, and they’re used in electronics manufacture. We wouldn’t be able to manufacture a billion-plus smartphones a year without them.

So you can imagine pairing a RepRap printer with a pick-and-place robot, and then you’d have everything you need to have a printer print a copy and assemble that copy into a new printer.

Of course you’d want to print a new pick-and-place robot at the same time as the 3D printer, so that the new 3D printer could both print and assemble itself.

So we’ll take this 3D printer and 3D pick-and-place robot and bundle them together — they’re now one machine, made of two smaller machines.

As it turns out there is a commons-based guide to building a simple pick-and-place robot using an Arduino, a commons-based hardware platform.

So we can continue to use commons based techniques of sharing and amplification throughout this process.

Let’s say we get all of that done, what have we got?

Well, we’ve got a machine that can replicate itself. Which is quite a thing.

This isn’t perfectly self-contained. Just as a living thing needs food and vitamins to replicate itself, this machine will always need some external resources — various metal bits, and lots of electronic components. Plus raw plastic.

But now that we have a machine that can replicate itself, it’s now possible to distribute that machine as widely as needed, because it can make all of the copies required, given sufficient resourcing.

Replicators! Everywhere!

So let’s say that there is not one of these machines, but — potentially — billions. So we can billions of machines that can make copies of themselves.

That doesn’t seem immediately useful — until you consider that everything this machine does is a function of the software running on it.

To change what this machine manufactures, change the software running on it.

Since the entire machine is commons-based, all of the software is open source, and can be freely inspected and modified by anyone to any purpose.

Someone could write a program to get the machine to assemble something else. Perhaps the folks at the Centre for Field Robotics write a program to get the machine to assemble a farming robot.

What they’ll learn quickly is that the machine is good at making copies of itself and less good at making farming robots. So they’ll modify the design to be a bit more general purpose — fit for replication and making farming robots — and share those designs with the commons.

Now folks can build machines that can replicate and build farming robots.

It isn’t very long before these additions to capacity become a virtuous cycle. That is, the more useful this machine becomes, the more people use it, learn from it, modify it, then share what they’ve learned.

The whole system undergoes a very rapid evolution from a very specific replicating machine, into a much more general purpose manufacturing machine, one capable of producing a broad array of useful devices — anything from lamps to farming robots to whatever other bits of kit people commonly need.

Again, all of this is commons-based. As soon as anyone shares a plan for a lamp or a chair or a robot, everyone with one of these machines has the same capacity.

Just as when someone adds a bit of knowledge to Wikipedia, it’s simultaneously available to more than a billion people, anyone who adds capacity to this commons-based manufacturing system brings it to everyone who has that system.

So the value of the system skyrockets exponentially as it becomes more valuable. Which brings more people to the system, designing more things, making it even more valuable.

This machine quickly becomes a factory for almost all of the kit anyone will need, from surgical equipment to toys, and it brings that factory capacity to every corner of the planet.

As always there is going to be a question of access to the input resources — some of those will be rare and expensive.

As this is a commons-based approach, every shortcut anyone finds for any input — substitutions that work as well or better, but more accessible — those will also spread quickly and ubiquitously.

The network of machines will quickly optimise its productive capacity, providing the greatest flexibility and material capacity for the users of those machines.

So the resourcing required to support a global-scale, decentralised, commons-based replicating manufacturing platform is not a fundamental drawback to its success.

What is? Inertia.

We are mired in ways of thinking about production and materiality that are old. That come from a time before a decade ago.

Now that everyone can potentially know almost anything, and now that we can put that knowledge to work in capacity, the only thing holding us back is our expectation that the future will be anything like the past.

A virtuous circle of sharing and improvements drive exponential capacity building.

But it won’t. We’ve already broken conclusively with the historic process surrounding knowledge formation. For the last twenty years knowledge has been collective and connective. Knowledge is the capacity amplifier.

So now we see one aspect of what that capacity amplification looks like.

And again, it doesn’t matter at all what we do here in Australia or in Silicon Valley.

This is commons-based work, so in Dhakka or Mumbai or Nairobi — where there are plenty of smart people, enjoying the same advantages of connected intelligence, they can work these things out.

This is the kind of project well suited national universities that want flagship projects that have real impacts — and it won’t cost a lot of money to fund. Commons-based processes are cheap and accessible.

This transformation in manufacturing is both post-Capitalist and post-Colonial, in that it generates a means of production no longer bound up in alienation. Quite the opposite: it is inherently participatory.

And that is precisely what makes it irresistible.

We’ve seen it in the rich West with the ‘maker’ movement — as soon as these tools became available, makers began to learn from them and improve them, sharing those improvements.

That tools revolution has now reached an exponential inflection point because of the concatenation of smartphones and robots.

In less than three years, eighty percent of all adults on Earth will be using smartphones — precisely because of the capacities they bring.

Connecting smartphones to robots is the natural next step — as is visible in the work of the Australian Centre for Field Robotics.

It makes both robots and smartphones more valuable, and — because the smartphone is pre-eminently the tool of shared knowledge — provides a framework for sharing all of the innovations that come from this increase in capacity.

We’ve got everything we need already. We don’t need to be optimists to see that. We only need to share what we know how to do.