Can the simulated discover the simulator?

One of the more outlandish proposals I discussed in Where is Everybody? was the “planetarium hypothesis” – an idea put forward by Stephen Baxter, one of the world’s foremost science fiction writers. Baxter argued that one popular idea in science fiction – that we live inside an artificial reality constructed by beings far in advance of ourselves – could actually be tested by experiment: even the most advanced civilisation would have constraints placed upon it by the laws of physics, and these constraints could be tested by experiment. If it turned out that we were living inside an artificial reality, that we were mere simulations in a computer program constructed by some mighty alien intelligence … well, that would be the Fermi paradox solved!

I didn’t take the planetarium hypothesis seriously, of course, but some people do. Nick Bostrom, for example, has taken the idea further. Bostrom is a professor of philosophy at Oxford University, and he has done some serious analysis of the notion that we are living in a computer simulation. Bostrom’s work is real philosophy, and mind-bending stuff. I urge you to visit his website, and think about his ideas: they are fascinating.

Three physicists who have clearly been influenced by Bostrom’s work are Silas Beane, Zohreh Davoudi and Martin Savage and last week they published a paper on arXiv (entitled “Constraints on the universe as a numerical simulation“) that examines some aspects of this seemingly bizarre notion. Beane, Davoudi and Savage are lattice gauge theorists, which means that they create a simulated toy “universe” in order to study quantum chromodynamics, the fundamental force that governs the interactions of quarks and gluons. They’ve extrapolated the  current trends in computational requirements for lattice QCD, and examined the notion that our own universe is a numerical simulation on a spacetime lattice. They then ask the question: would there be any observable consequences if we were living in such a simulation?

Read their paper for the details. But their final sentence sums it up: “assuming that the universe is finite and therefore the resources of potential simulators are finite, then a volume containing a simulation will be finite and a lattice spacing must be non-zero, and therefore in principle there always remains the possibility for the simulated to discover the simulators.”

Stephen Baxter showed how a K3 civilisation could create a perfect simulation with a radius of about 100AU. Well, Voyager 1 has already passed beyond that distance (as I write, it’s at a distance of 122.64AU) and it didn’t bump into a metal wall painted black. So we know – by experiment! – that we don’t live in a K3 civilisation’s perfect simulation (though of course the Voyager result hasn’t ruled out the possibility that we live in an imperfect simulation; the boundary of an imperfect simulation can be much further away). The work by Beane, Davoudi and Savage provides us with other tools for testing whether we inhabit a simulation.

And just for the record: no, I don’t think we live in a simulation!