“But where is everybody?”
Enrico Fermi
Can a machine replicate itself? Given enough construction materials, such a robot could multiply exponentially — one robot copies itself into another robot, two robots become two, four robots become eight, and so on.
Potential applications for a self-replicating robot are humbling. Imagine an unmanned space probe that could explore the galaxy while also harvesting raw materials from passing asteroids and planets to build copies of itself. We would only need a few dozen generations of these probes before we could assign a single probe to each of the 250 billion stars in the Milky Way, and in so doing explore every planet in our galaxy for signs of life. Such probes are called von Neumann probes in honor of our friend John von Neumann from our work with pseudorandom number generation.
Many scientists believe that a technologically advanced society could produce von Neumann probes. They also believe that the conditions for life almost certainly exist elsewhere in our galaxy. Why, then, have we not seen evidence of alien life ourselves? This conundrum, inspired by Enrico Fermi’s alleged quotation above, is known as the Fermi paradox.
Self-replicating robots are difficult to build (see video below). Despite major advances in robotics research, we are still far from manufacturing a robot that can both make a copy of itself and perform a practical task competently.
To understand self-replication, perhaps we should look for the simplest self-replicating system that we can find. This was one of John von Neumann’s goals in the 1940s — which is why hypothesized self-replicating space probes are named in his honor. When Stanislaw Ulam pointed out to von Neumann that cells, the basic unit of all living things, are themselves miraculous self-replicators, the strange and wonderful world of cellular automata was born.
