Skip to main content

Von Neumann Probe

It's hard to create a list of all the scientists in the history of mathematics and physics who better applied theoretical knowledge into the hardware that ultimately worked and moved the world to the next step of existence. But if we try to do it, John von Neumann would be among the top five in the list of scientific GOATs. Probably the best of all of his contributions is in the history of computer science - along with Arthur Burks and Herman Goldstine, he published a paper* in 1946 that practically described the architecture of a modern computers as we are familiar with today.


However, what he will be most remembered for is not a machine in existence today but one that is still just a theoretical and basically only an idea. Not yet anyways. In short, design of a von Neumann probe or a self-replicating spacecraft is not that far-fetched from all what we know today and if humanity sees its survival on Earth difficult in the future (to say the least) and tries to became interstellar or at least interplanetary species, von Neumann probe will indeed become the reality. If realized, that invention could, one day, be used to explore entire Milky Way galaxy in a relatively small amount of time.

The idea behind the probe is to use materials already available and accessible on rocky planets, moons and asteroids. The self-replicating spacecraft without all the limitations introduced by humans aboard (food, energy, small acceleration, sudden changes of direction, artificial gravity, health hazards, etc) would be able to travel from system to system, mine all the materials it needs to print even more spacecrafts and probes and move on. Ultimately, reaching the exponential curve of self-creation, the single craft would be able to build a vast number of copies of itself traveling in all directions.


By the worst estimate**, it would take up to 10 million years to dispatch millions of probes across the Milky Way (assuming the travel speed is about 1/10th of the speed of light). Of course, the first goal of exploring our galactical neighborhood, say within 100 light years in diameter, would be probably measured by centuries instead.

There are numerous challenges behind the machine exploration in such fashion and probably the main one would be to not have exponential self-creation be the goal itself and instead to send probes into real missions. For example, what come first to mind is to set the goal to explore nearby systems only for finding the nearest Earth-like planet and stop as soon as the mission is achieved. The replicated probes could be given new assignments after, i.e. to serve as ice haulers from outer system to inner planets if needed or to serve as a science probes from different regions of the planetary system they were last located.


I wanted this blog post to be both, short description of von Neumann's vision being one of Jules Verne's type of potential future and a small recommendation for one amazing novel series. In the science fiction, there are lots of 'von Neumann probe' plots in all media, especially acting as berserkers or villains in various novels and movies such as the Borg of Star Trek, a self-replicating bio-mechanical race or insect-like replicators from Stargate SG-1 TV show. 

Not all of them are bad guys and a trilogy I mentioned was 'We Are Legion (We Are Bob)' by Dennis E. Taylor. What brought my attention to read it was the premise of not using the AI as a pure software entity and Dennis introduced instead an AI originated from a real human brain scan. In the series, Bob Johansson wakes up a hundred years later after his death only to realize to become a perfect computer emulation of himself with all of his memories preserved. Bob's personality is what adds another layer of 'humanity' to the entire endeavor and I warmly recommend the books and all the adventures in the Bobiverse. 

Image Refs:
https://www.theodysseyonline.com/luxembourg-plan-asteroid-mining
https://www.amazon.com/gp/product/B0753LBFQ7/ebooks


© 2023 Milan's Public Journal