ChipLog

Various stuff from some guy.

Universal segfault

12 Comments Posted by chipx86 on April 21, 2007

(I’m sick today with a stomach flu and have recently consumed NyQuil, so this may make much more sense in my own head. I’m also not an expert in quantum mechanics, so please excuse any misconceptions I have.)

I was reading an article about quantum mechanics and how many scientists in the field want to find a definite reasoning for the whole “Spooky action at a distance,” where two particles can be quantumly entangled, causing changes to one particle to affect the other, regardless of the distance between them.

So I began to think about this. I imagined the universe as a simulation program, exposing basic building blocks into the simulator that everything was built upon, so tiny in comparison to anything the developer would expect to grow or evolve in the simulator. The developer powers it on, watches for a bit, goes to grab a cup of coffee, and checks back periodically. Perhaps he even interferes just a bit in the affairs of the simulated life inside the system, and then goes away for a while.

Now at some point the “life” in the system gets curious and starts heavily investigating those basic building blocks. So far the simulation has been pretty stable, but now the basic building blocks are being messed with, poked at, accessed in ways that weren’t intended. And like any software, there’s bugs. The lifeforms in the system manage to get two of these objects to entangle with each other. They marvel at the fact that the objects are still somehow linked even across great distances. To the developer, this wouldn’t have been that surprising. Afterall, they’re just two objects with various properties, and location is just one of them. Perhaps they’re even designed to share certain types of data for memory optimization reasons, and they don’t sanity check locations between linked objects for performance reasons. It may not matter that much to the developer, but it certainly managed to confuse the lifeforms in the system poking at the objects. Fortunately it appears safe, but they’re poking at a whole lot of different things that weren’t meant to be exposed, and no program is bug-free.

So pretend for a second that this was the case for our universe.

What if it crashed?

Uncategorized Food-for-Thought, Life, Quantum-Mechanics, Universe

← Netvibes Universe VMware Workstation 6 RC2 →

12 Responses to Universal segfault

randomwalker April 22, 2007 at 12:27 AM
Here’s a related question: what if the programmer did

killall -STOP simulator

went to get coffee, came back, and did

killall -CONT simulator

What would it look like from our point of view? A minute of thought should convince you that we would notice nothing different at all. Similarly, if it received a SEGV, our universe would just cease to exist, but not with some kind of bang; we would simply not be able to perceive anything anymore.

A guy wrote a whole novel about this stuff, called permutation city. One of the threads there is a guy simulating himself in a universe and messing with the simulator in all kinds of interesting ways.

Reply
Matthijs van der Vleuten April 22, 2007 at 12:31 AM
If the simulator crashed, it would simply stop running.

Reply
Adam April 22, 2007 at 12:33 AM
Heh. Interesting thought. Have you seen http://www.simulation-argument.com/computer.pdf ?

Reply
Daniel April 22, 2007 at 12:49 AM
Funny, I don’t even need NyQuil to go that far. I blame videogames

I thought of basically the same idea a couple of days ago (while taking a shower), but in my case the system had a purpose other than simulation (storage? computation?) and the universe (or thermodynamics) was pretty much the bug. But I’m sure it wasn’t an original idea (in my case), just a mash-up of weird ideas I’ve heard or read about.

What I find funny thing is that I’m an atheist biologist and that idea is sort of like Intelligent (but Fallible) Design. I don’t consider the whole thing probable or even feasible, but I know that it’s something we’ve been exposed to for a long time (think Chariot of the Gods, 2001, Matrix, The Bible, XFiles, HHGTTG, etc.). It’s interesting to discuss what we come up with both as a mental exercise as to gauge the influence of these memes on us… I guess it might indicate the expected outcome from these influences would be “gee, [Infallible] Intelligent Design makes a lot of sense”. Perhaps that’d be expected without any influences, I don’t believe it but it might be so

If you find yourself with time to spare, check http://outoforder.adventuredevelopers.com/ for a IMHO nice game that is based on a… related scenario. There is a walkthrough available online, in case you get impatient like I did

Reply
harrytuttle April 22, 2007 at 1:45 AM
“What if it crashed?”

well the real question would be: what are they trying to achieve with the simulation?

but i guess they could just they pick the latest backup (hoping that’s not in the middle ages or before…), debug the crash, fix it, and start again. and there will be a universe without quantum mechanics…

another interesting question is: could this have already happened in the past? and could we guess when?

Reply
Torsten April 22, 2007 at 2:24 AM
I had the following suspicion for a long time:

What is up this this how quantum things: Why is there a minimum energy quantum that can be transfered ( http://en.wikipedia.org/wiki/Planck's_constant )? Why is even the position of subatomic particles quantimized ( http://en.wikipedia.org/wiki/Planck_length )?

This just looks like the simulation is running on integer math and the Heisenberg uncertainty is the end of its resolution.

Reply
Murray Cumming April 22, 2007 at 6:55 AM
It’s not a bug. It’s just poorly documented.

Reply
Dan April 22, 2007 at 8:54 AM
Truly a software developer’s point of view on QM
What if the opposite is true.. suppose you and your computer are experiments carried out by sentient quantum creatures. What if they get bored?

Reply
Aaron April 22, 2007 at 11:32 AM
Hopefully, if the simulator crashed the programmer would fix the bug, rewind to a point just before the failed experiment, and restart it.

Reply
Johan April 22, 2007 at 2:37 PM
People have thought about similar things – Gerard ‘t Hooft has tried to develop some sort of cellular automaton model for quantum mechanics. Few people take those attempts seriously, for good reason. Among other things, they seem to violate Bell’s theorem.

Reply
Tony April 23, 2007 at 9:12 PM
Any developer capable of designing something as heinously complex as this universe wouldn’t get tripped up by a bad dereference or an overflow. I’d be more worried about deadlocks, personally.

Reply
Niklas April 26, 2007 at 11:04 AM
Deadlocks. You mean doing something funny might make time stop in parts of our universe. That could bee interesting!

Reply