The Simulation Hypothesis Part 1 – Nick Bostrom’s ‘Simulation Argument’

 The Simulation Hypothesis

 

reality-media-Alice-matrix-lo

All this talk of the future brings us nicely to the simulation hypothesis. This is the idea that we live in a form of virtual reality. That our universe and experience of space time is a simulation run from outside space time (extrinsic simulation) or from within it (intrinsic simulation).  Odd though it may seem, it’s hard to disprove and in fact seems to support much of what we see in physics. This is an overview of the current thinking on the subject along with my own meanderings.

The simulation argument


Nick Bostrom’s paper
in 2003 introduced the simulation argument. A logical speculation on our future possibilities with virtual reality technology. He refers to the idea of ‘ancestor simulations’. Completely immersive virtual realities based on the world and evolutionary history, familiar to the beings that create them. The argument assumes that a post-human civilisation of the future would have enormous computing power. This seems reasonable when projecting indefinitely into the future given the exponential growth in computer technology today. It also assumes substrate independence. This is the idea in philosophy jargon that ‘mental states can supervene on any of a broad class of physical substrates’. In other words, consciousness is not confined to carbon based neural networks like the brain but can also be implemented in other forms like a suitably powerful computer. This is a controversial assumption that opens up discussion on the nature of consciousness. He describes a post-human civilisation where we have merged with technology and have enormous computer power and intelligence. Using temporal Logic, he then argues that at least one of these three possibilities MUST be true.
1. Human level civilisations never reach this post-human technologically mature state because they go extinct before reaching it.
2. Post-human technologically mature civilisations do exist though they lose interest in developing ancestor simulations for whatever reason.
3. We are most likely in such a simulation.

• Possibility number one is of immediate concern. Averting extinction in the midst of the existential risks posed by our advancing technology seems like a long shot. Molecular nanotechnology is one of many likely candidates among existential risk speculators. A kind of mechanical bacteria that developed for malicious ends or by accident could very quickly wipeout most life on the planet. Unless our psychological and ethical evolution can keep pace with the technology we’ll be hard pushed to make the right decisions in the precarious world of the future. For the purposes of the argument though, possibility 1 doesn’t need to be quite so catastrophic. Civilisation would just need to collapse, locking humanity into a continual primitive state, preventing it from reaching technically advanced post-humanity.  This would also assume that most human level civilisations in existence would likely evolve the same way.

• Possibility 2 assumes that we will decide not to run ancestors simulations and that all intelligent civilisations in existence would likely converge in this way. Maybe due to ethical considerations, law or simply that our post-human descendants have no interest in it. It will be interesting over the coming decades to see if our desire to create virtual reality experiences continues.

• If we ever develop a convincing virtual reality simulation and still have the desire to run it, the chances that number 3 is correct would increase. We would likely already be in a simulation ourselves. The converse is also true. If we’re not in a simulation it’s unlikely that our descendants will ever run one. If we do reach the point of creating a simulation we can safely assume we are in one ourselves and that the creators of it were also simulated. Nested simulations created within other simulations. Some have likened this to virtual computing that we see today. Java script and web-applets for instance, run on a simulated computer inside a physical one. Nick Bostrom then goes on to postulate on the amount of computing power required to build a sufficiently convincing simulation. Given our lack of knowledge at present he considers it reasonable to attribute even proportions to the three possibilities.

God?


reality media godthefather

This starts to sound something like the idea of God. The philosopher David Pearce said in relation to Nick Bostrum’s paper
The simulation argument is perhaps the first interesting argument for the existence of a creator in 2000 years
An ancestor simulation in the style of Nick Bostrom’s simulation argument would require a creator. Presumably one nostalgic for the world it knows and recognises. In this case, if we are in a simulation ourselves we might assume that the creator programmed evolution according to the evolutionary laws by which it itself knew and evolved by. Such a creator is in accordance with the regular view of evolution and requires no existence of an eternal God. This seems like a more palatable idea to atheists. It seems equally possible though that the creator of a simulation would have no likeness to us whatsoever. A simulation could be created and then evolve according to its own intrinsic laws. There needs to be no mastermind overseeing it as it progresses. Although the simulation hypothesis suggests a creator it does nothing to speculate on the starting point of consciousness. Consciousness would still require an initial material universe in which to evolve if the assumption is that consciousness has indeed evolved from a pre-existing material universe. Another perspective is that consciousness is not a product of an external universe but is in fact the creator of it. More on that later.

On the 3 possibilities laid out in the simulation argument.

It seems to me that in order to survive and navigate the complex and potentially dangerous technology of the future, we would need to develop a consciousness very different from that which we have at present. We would need a much greater understanding of reality and be able to act as a cohesive group. With greater cohesion among sentient beings comes empathy and ethical maturity. Without this, extinction seems certain. I can’t envisage a midway point where advanced, intelligent civilisations race around the galaxy fighting each other. I’d expect future technology to advance so fast that conflict would result in extinction very quickly. So assuming we do evolve into some kind of enlightened civilisation, tempering conflict and attaining a high level of foresight and awareness, would we run such a simulation if capable of doing so? I can’t see any advanced post-human civilisation having any desire to maintain a Truman Show style of virtual reality on sentient beings. I think the idea of maintaining and overseeing a simulation would be considered pointless by any civilisation intelligent enough to survive that far into technological maturity. I could however see an interest in cultivating life and consciousness. If at some point in the distant future we manage to program a universe out of some kind of big bang and evolve life and consciousness according to some predetermined set of rules could this be considered a simulation? This might not require the astronomical computing power of maintaining a simulation as outlined in Nick Bostrom’s argument. A relatively simple set of rules could then potentially create an intricate universe like that which we see around us. A simple example might be Conway’s game of life.  A cellular automation mathematical game whereby ‘it’s evolution is determined by it’s initial state, requiring no further input’. There are only a few initial rules though the possibilities are huge, creating highly intricate patterns. If a universe was creatable by some future technology then we’d presumably go on to create many. We would most likely be in one of these ourselves. It seems feasible to me that the underlying physics of the universe could have been programmed by a consciousness such as our future selves. If we were able to ‘grow’ sentient life by creating a material universe, each universe would then spawn more universes ad infinitum. Future speculations aside, there is a growing school of thought that our own consciousness does in some way creates our reality and perception of a material universe. From this perspective all reality might be considered virtual. This is considered in the next post.

Leave a reply