Talk:Building the Simulation

From Simulism

This is a nice effort. But should we begin from scratch or use some existing virtual world (a la Second Life) as the basis? --Ivo 11:27, 15 April 2007 (CEST)

[edit] Reasoning for Ground Up Approach

Thanks for bringing up the issue of adapting an existing simulation. If it were possible to bend what someone else has already created for these purposes that would be an acceptable start for building a simulator. Unfortunately 'Second Life' and other online simulations are not much more than enhanced online communities. They don't allow you to test theories or change the underlying processing of the simulation and aren't exactly free and open either.

If you look at what most simulators are doing today, they focus like a laser beam on their area of interest and ignore everything else around them. Unfortunately, this criples their usefullness as tools for research across many disciplines. They essentially create a one dimensional cross-section of the world and they keep their knowledge and progress to themselves. In an ideal simulation framework, a company working on a drug might submit a compound to the repository, another company may be trolling through the database searching for a material that will work well for their application and find company A's compound works for their application. You could even credit the creator of the entry.

I envision a two tiered system where you have one simulation constantly running where you could create things to test with the accepted best version of physics and science and another where you could test your own modifications to the engine if you so desired. You could run a subsim, if I can coin the phrase, that allows you to examine only what you are interested in while optimizing out all the other variables and change those parameters like a sort of meta-zoom, which is not possible at this point with any programs I know about.

I don't think the avitar based simulations are capable of building something so dynamic because they were not built from the ground up for that purpose. What do you think? Ethicalhacker 01:07, 17 April 2007 (CEST)

[edit] Alternative to Planet-sized Computer

Would a computer the size of a typical terrestrial planet be enough, especially if those running the simulation had multiple simulated worlds/universes running at the same time? It seems much more likely that if you can run one simulation you will want to run many more. Every physicist, anthropologist, sociologist, biologist and political scientist will want one, and that's just for the sciences. When you get into entertainment simulations... well, just look at how many web pages there are on the net now, and the rate of proliferation over the last ten years.

Therefore, I think a planet computer wouldn't be enough. I think they'd be using brown dwarf stars. For one thing, brown dwarfs above 13 Jupiter masses are able to fuse deuterium, and those above 65 Jupiter masses are able to fuse lithium. This would provide a power source. Secondly, with masses of this magnitude they would have all the raw material needed for construction of computing mass of whatever size required. Thirdly, with so much mass the simulators might begin with only a small fraction of the mass of the brown dwarf and have material in reserve for later construction as demand required without going off-site. Fourth, if you begin construction in the middle depths of the dwarf you would still have sufficient atmosphere to safeguard against catastrophic meteor or asteroid strike, which might be a problem with a mass that large.

Using a brown dwarf would also provide useful camoflage against nosy lesser beings and their telescopes, Any extra heat and mass generated would be assumed to be the natural properties of the dwarf and not the evidence of any higher intelligence at work.

--AuntyProton 16:25, 6 November 2007 (CET)