As part of the reading I'm doing for my
GCIH re-certification, I've been reading about software that tries to detect whether it's running on a real machine or on a virtual machine. This makes me think about how a person might try to detect whether the world is real or virtual, not philosophically but in practical terms assuming some limitations on the part of the software.
There are 4 common techniques used today by malicious software to detect virtual machines and I'm going to try to restate them as though they applied to virtual worlds:
- Look for artifacts in the physical environment (in software, they refer to processes, file system, or operating system registry)
- Look for artifacts in the world's awareness of itself (in software, they refer to memory, but I think for this proposition we could mean linkages between separate "physical" things that either should be there and aren't or vice versa)
- Look for physical things that are clearly identified as only belonging to virtual environments (in software, they refer to virtual hardware, like virtual network cards that always have the same set of MAC addresses)
- Look for virtual environment specific features or capabilities that don't work in the real environment (in software, they refer to processor instructions that only work in the virtual machine or that work a specific way in the virtual machine)
In the Matrix, 3 of these things are clearly present to tell you that the environment is virtual. The reference to 'deja vu' in the movie is clearly an artifact of the Matrix. The agents are clearly only present in virtual environments. And the abilities of both agents and resistance fighters within the environment would clearly not exist outside of that environment. Which is fine in terms of thinking about the Matrix, but the resistance knows they're going into the Matrix each time and essentially has exploit code that bootstraps them when they connect. Ordinary people are indoctrinated or reprogrammed or removed as necessary, but these guys always start with elevated privileges.
This led me to think about a group of people whose job is to hop through virtual worlds. They're indoctrinated during loading each time and must work out from scratch that the world is virtual and how to hack the world. So a key factor in the story is not only how to tell that it's virtual, but providing training that will survive indoctrination and give them the tools they need. The necessary assumption is that at least some virtual worlds are not consensual places. And they don't start out with elevated privileges. They need to hack from the inside to get those.
[Edited to add:] The more I think about it, the more it seems like you'd want hard science folks looking at things from the inside. Looking for artifacts seems on the face of it to be one of the easiest ways, even if it's something goofy like rounding in some physical constant or rounding errors that crop up in certain interactions. What if quantum entanglement (or a lack of it) were a sign that the environment was virtual? Would we be able to detect that the "universe" was losing seconds periodically using an atomic clock? (That might be a sign that we were running against another universe's clock.) What biological processes might be simplified at the molecular level or higher up (and how would it be detected)?