Science Forums 2

[CraigD]

Originally Posted by Solve

It's amazing how little has been posted on Nick Bostrom and the simulation argument!

Indeed. How could we have all missed this some 1221 days ago?

The “simulation hypothesis” has been discussed a few times around these forums, but by no means, IMHO, to exhaustion.

Originally Posted by Alma-Tadema

One thing that always bothered me is the computing power required to run a meaningfull simulation of an universe
...

One-time poster Alma-Tadema voices a lot of intuitive objections to the hypothesis, most of which most of us have entertain at various times. I think some application or methodic reason over intuition can answer many of them, though...

The problem with parallel computers is the communication delay between nodes, the speed of light limit

Along with the “if you try to get enough computing machinery close enough together to not have long communication delays, it’s Schwarzschild radius will exceed its radius, and it’ll become a black hole” issue, this seems a pretty inescapable limit. However, this reasoning appears to me to err in assuming that a reality simulation must run in realtime relative to some external reality. As with the ordinary simulations some of us have written and run, this isn’t the case. A reality simulation could be running much more slowly than the external reality of its computer, with no effect on the simulation. Because, according to the simulation hypothesis, everything we simulated beings can use to measure time is part of the simulation, we’d be unable by any means detect the ratio of simulated to actual time. Thus, a computer with synchronous parts separated by thousands of light-years would have an effective clock speed of on the order of .00000000001 () Hz, but this intuitively ridiculously slow rate would be insignificant to the simulated reality.

Such machines would, however, require vast durations to simulate even a single minute, day, or human lifetime, easily exceeding the duration of the Stelliferous era (about to years ABB, at which we’re at nearly the very beginning, at years ABB. As essentially all of our present day technology is directly or indirectly star-powered, such a computer would have to be powered differently. Fortunately (for our imagined very far-future simulation makers), the post-stelliferous universal eras are expected to have lots of even greater power supplies, primarily super-massive black holes.

One way that negates all that is if you can find a way to access an infinite pool of processing resources. Now the only way that we currently know of is Prof. Tipler's Omega Computer hypothesis.

This is referring to an idea by Frank J Tipler, described at length in such books as the The Physics of Immortality. Tippler’s hypothesis – some believe it’s more an expression of desperately wishful thinking than a serious scientific hypothesis – has several requirements not required by the simulation hypothesis

• The simulation must match “heaven” as imagined by every being that has ever lived and imagined it.
• The simulation must be latterly eternal – that is, it requires a actually infinite amount of processing.
• At present, as we’re not in heaven, we’re in the same, real universe in which the omega computer will someday be.
• Somewhat as an aside, and requiring a fairly lengthy explanation found in the TPoI, the physical universe must be closed (). Tipler appears to suspect that the universe is not naturally so, but can be made so artificially.

Without these requirements, a reality simulation per the simulation hypothesis is quite literally infinitely easier than with them. At the same time, Tipler’s omega point idea explicitly rejects one of the more intriguing assertions of Bostrom’s simulation hypothesis: that it’s likely we are, right now, in a simulated reality.

Both ideas seem to me to share the assumption that the presumed current or ultimate far future reality simulation is artificial, an assumption I think is somewhat limiting, as it’s not to difficult to imagine that actual, physical reality is a sort of simulation run on an underlying “hardware” consisting of physical reality itself.

So my conclusion is that the simulation argument by Bostrom is right only if Tipler is right or you can access an infinite amount of processing power(only Tipler has currently shown how this may be possible).

My take on these two’s ideas comes to nearly the opposite conclusion. As I noted above, it appears to me much easier for Bostrom’s hypothesis to be right than Tipler’s.

Originally Posted by Southtown

Life cannot simulate itself anymore that generators can surpass input with output.

I don’t see how this analogy follows.

Electric generators can’t produce more output than input energy due to the law of conservation of energy. There’s no requirement that a computer simulations violate this law. In obvious fact, every one yet written follows the usual physical laws, and, barring the sort of far future super-science hinted at by Tipler, will almost certainly continue to do so.

Although progress has been slow – surprisingly to many biologists and computer scientists – it’s not unreasonable, I think, to expect that projects to accurately simulate simple life such as the E-Cell project, will eventually succeed, demonstrating that life, in the form of human scientists, truly can simulate life.