In a 2003 article, Nick Bostrom, a philosopher from Oxford University, outlined the possibility that our reality is a computer simulation invented by a highly developed civilization. In this paper he argued that at least one of the three sentences must be true: civilizations usually die before they have developed the ability to create simulations of reality; developed civilizations are usually not interested in creating simulations of reality; we almost certainly live inside a computer simulation. Recently, Columbia University astronomer David Kipping scrutinized these provisions, also known as Bostrom’s “trilemma”, and proved that the probability that we are actually living in a simulation is 50/50, reports Scientific American.
Creatures inside a computer
So, let’s imagine that we are virtual beings living in a computer simulation. If so, the simulation is likely to create a perception of reality on demand, rather than simulate all reality all the time – just like a video game optimized to display only parts of the scene visible to the player. Astrophysicist and science popularizer Neil Degrass Tyson in the talk show “Star Talk” (Startalk) suggested that maybe that’s why we can’t travel faster than the speed of light.
Of course, such conversations may seem frivolous. But since Nick Bostrom wrote the seminal article on simulation, philosophers, physicists, technologists, and even the average citizen have been trying to figure out how we can figure out whether we live in a simulation or not. I note that most researchers are skeptical about the idea of the virtual universe, but Kipping’s work shows that if people ever developed the ability to imitate conscious life, they would most likely be creatures inside a computer.
In 2003, Nick Bostrom imagined a technologically advanced civilization that has enormous computational power and needs some of that power to model new realities with conscious beings in them. Given this scenario, his simulation argument showed that at least one sentence in the next trilemma must be true:
-first, people almost always die before reaching the modeling stage;
-Second, even if people reach this stage, they are unlikely to be interested in modeling their own past;
-and thirdly, the probability that we live in a simulation is close to one.
To better understand Bostrom’s argument about the simulation, Kipping resorted to Bayesian reasoning. This type of analysis uses the Bayesian theorem, named after Thomas Bayes, a British mathematician who lived in the 18th century. The Bayesian analysis calculates the probability that something will happen (the so-called “a posteriori” probability) by making assumptions about the event under analysis (assigning an “a priori” probability to it).
Do we not live inside the simulation?
Kipping started by turning the trilemma into a dilemma. He summed up the first two sentences into one, claiming that they both would lead to the same result – we don’t live inside the simulation.
We simply assign an a priori probability to each of these models and proceed from the principle of indifference, which is the default assumption when you don’t have any data or proclivities,” Kipping said in an interview with Scientific American.
Kipping also claims that the more layers of reality are built into the simulation (as in a nesting doll), the less computer resources are required. In other words, the further you go down the rabbit hole, the less computing power you’ll need to create a compelling simulation. The astronomer’s conclusion after recalculating the numbers was this: the probability that any of the hypotheses is correct is about 50 percent.
But if people ever came up with such a simulation, the picture would change drastically. The next stage of analysis required comprehension of “parody” realities – those that can generate other realities – and “unrelated” realities – those that cannot. If the physical hypothesis were correct, the probability that we live in a barren universe would be easy to calculate: it would be 100 percent.
Kipping then showed that even in the simulation hypothesis, most of the simulated realities would be barren. This is because as the simulations generate new simulations, the computational resources available to each subsequent generation are reduced to the point where the vast majority of realities will be those that do not have the computational power to simulate future realities capable of accommodating conscious beings.
Put it all together in a Bayesian formula, and you get the answer: the a posteriori probability that we live in a basic reality is almost the same as the a posteriori probability that our world is a simulation.
However, if the simulation has infinite computing power, then under no circumstances will we see that we’re living in a virtual reality, because it can calculate everything we want with the degree of realism that we want. But if we find that we’re living in a simulation, that means it’s likely to have limited computing resources. What do you think is real and can it really be a simulation?