Are we just a simulation in simulation

Are we living in a simulation?

Is everything just virtual? In science fiction works like “Matrix” or “Simulacron” we all only live in a simulation. Researchers have now investigated whether this is technically feasible and likely. Their result: The effort to convincingly simulate so many intelligent beings and their intellectual performance and environment would be gigantic even for a highly advanced civilization - and the result would not be flawless.

How real are our world, the cosmos and everything in it? Is our universe just a holography - a projection of two-dimensional fields and parameters? Or maybe we even live in the computer simulation of a cognitively and technically far superior civilization? People have been debating this question for centuries, and there are countless variations of this scenario in science fiction.

The philosopher Nick Bostrom sees it as almost inevitable that an advanced civilization will create simulated worlds - unless it self-destructs before it reaches the necessary technological capabilities. At a conference a year, SpaceX founder Elon Musk put the chance that we are real at just one in a billion.

How likely would a simulated cosmos be?

Alexandre Bibeau-Delisle and Gilles Brassard from the University of Montreal have now carried out a new probability estimate. Unlike earlier approaches, in their mathematical model they also investigate the possibility that our thought processes and interactions with the environment are simulated using quantum computers and quantum operations.

"Like many things in computer science, the idea that our world could be a simulation needs to be reconsidered in the light of advances in quantum computing," the researchers state. “To want to reproduce all of our physics with classical resources alone hardly seems feasible.” They also check what amount of energy would be necessary and whether simulated beings could also carry out simulations on their part.

First requirement: simulate intelligent beings

A first important question would be how much computing power would be required to simulate the performance of the human brain on its own. The computing power of a brain weighing around 1.4 kilograms varies between 1014 and 1016 Operations per second, ”explain the researchers. Theoretically, assuming the appropriate technologies, computing capacities of up to 1050 Operations per seconds and kilograms.

If a civilization has advanced enough to convert at least a billionth of the energy stored in matter into computing power, then with just a computer the size of a brain they could develop 1.4 times 1025 simulate virtual brains, as the scientists explain.

Second requirement: interactions and laws of nature

But the virtual brains alone are not enough: A simulation must also record all interactions of the simulated beings with their environment and also all processes that result from the simulated laws of nature. "If you only want to create worlds in which intelligent beings live and interact, it would be advantageous to choose natural laws that allow intelligence but minimize the costs of simulating the environment," write Bibeau-Delisle and Brassard.

If a civilization wanted to create a virtual world whose physical laws correspond to those in reality, then the effort would be much greater: “It would be extremely time-consuming to then simulate this environment down to the smallest microscopic level,” explain the researchers. According to their calculations, the computing power for this would be so high that not even a maximum computing density would be sufficient to simulate a large number of individuals right down to these details.

However, it would be conceivable that the simulators adapt the complexity variably: If people are currently carrying out experiments that require a high level of detail, for example in quantum physics tests or astronomical explorations, then the "resolution" is increased. If our attention is not focused on such complex processes, the level of detail can be reduced.

The problem of recursive simulations

Another factor could increase the computational effort for a world simulation considerably: If the simulated beings in turn begin to use computers and carry out their own simulations. Then the computers of the real level not only have to maintain the simulation of the first level, but also their simulations - and so on. These recursive simulations would allow the necessary computing capacities to increase exponentially.

"It becomes even more dramatic when the simulated civilization uses large amounts of computing power for various other purposes - in addition to its own simulations," explain Bibeau-Delisle and Brassard. Then the computing power of real civilization might no longer be sufficient to simulate a large number in virtual individuals.

"In the absence of additional evidence, the number of simulated beings is therefore a good guideline for the probability that we will live in a simulation," said the researchers.

"Probability far below 50 percent"

Similar to the famous Drake equation, with which one can calculate the probability of the existence of extraterrestrial civilizations, all these factors can also be combined for a world simulation. With this calculation, Bibeau-Delisle and Brassard come to the conclusion: The chance that we will live in a purely virtual world is rather small and is in any case well below 50 percent.

"The main factors for this low probability are the enormous effort to convincingly simulate the environment of a civilization, the inevitably imperfect efficiency of every computer operation and the fact that simulations can be recursive," the scientists write.

Missing aliens are suspicious

However, there is one factor that could suggest that we are living in a simulation: the fact that we have still not discovered any extraterrestrial life or its probes. "If we live in a simulation with simplified physics, we will never encounter such probes," said Bibeau-Delisle and Brassard. Because such a simulation would save the effort of creating further beings far away from the simulated earth.

"The fact that we have not yet detected any evidence of extraterrestrial civilizations could therefore be seen as the most convincing argument for simulation theory," the researchers state. (Proceedings of the Royal Society A: Mathematical and Physical Sciences, 2021; doi: 10.1098 / rspa.2020.0658)

Source: Royal Society

March 8, 2021

- Nadja Podbregar