April 7, 2009

Welcome to the Machine, Part 1: The ethics of simulated beings

Without a doubt some of my favorite video games of all time have been those that involve simulations, including SimCity and The Sims.

When I play these games I fancy myself a demigod, managing and manipulating the slew of variables made available to me; with the click of a mouse I can alter the environment and adjust the nature of the simulated inhabitants themselves.

There's no question that these games are becoming evermore realistic and sophisticated. A few years ago, for example, a plug-in was developed for The Sims allowing the virtual inhabitants to entertain themselves by playing none other than SimCity itself. When I first heard about this I was struck with the vision of Russian Matrioshka nesting dolls, but instead of dolls I saw simulations within simulations within simulations.


And then I remembered good old Copernicus and his principle of mediocrity: We should never assume that our own particular place in space and time is somehow special or unique. Thinking of the simulation Matrioshka, I reflected on the possibility that we might be Sims ourselves: Why should we assume that we are at the primary level of reality?

Indeed, considering the radical potential for computing power in the decades to come, we may be residing somewhere deep within the Matrioshka.

Consequently, we are all faced with a myriad of existential, philosophical and ethical questions. If we are merely simulants, what does it mean to be alive? Are our lives somehow lessened or even devoid of meaning? Should we interact with the world and our fellow simulants differently than before we knew we were living in a simulation? How are we to devise moral and ethical codes of conduct?

In other words, how are we to live?

Well, there's no reason to get too excited over this. It's a bit of speculative metaphysics that doesn't really change anything -- assuming we are in a simulation, we should live virtually the same way as if we were living in the "real" world.

That is unless, of course, those running the simulation expect something from us. Which means we need to figure out what it is exactly we're supposed to do...

Tomorrow - Part 2: Descarte's 'Malicious Demons.'

8 comments:

Mac said...

I did a double-take when I saw the first picture in this post. I thought the Russian dolls were toes.

XiXiDu said...

What if a being becomes so complex that its theory of mind is as complex as a mere human being. Is thinking of harm then unethical?

At what point is the simulation of a being experiencing harm as unethical as the harm of what we call a "real" being today?

Tim Tyler said...

Hanson's take:

"If you might be living in a simulation then all else equal you should care less about others, live more for today, make your world look more likely to become rich, expect to and try more to particpate in pivotal events, be more entertaining and praiseworthy, and keep the famous people around you happier and more interested in you. "

- http://hanson.gmu.edu/lifeinsim.html

Duncan said...

"... Copernicus and his principle of mediocrity: We should never assume that our own particular place in space and time is somehow special or unique. ...: Why should we assume that we are at the primary level of reality?"

Or just the other way round: Living in a simulation is something special. Ergo: Why should we assume that we are at a simulated level?

Martin Andersen said...

Duncan: If there are a lot more simulated people than real people, then you are special if you are real.

Duncan said...

@Martin Anderson:
That there are indeed more simulated people than real people, is not conclusive in the pertinent literature, I think. And you say "If there are a lot more ..." -- yes: "if" -- but "if not ...?" -- or "if there are no simulations at all ...?. I think, nobody knows. I can see only unfounded speculations.

Above that, "not being particular" did *not* mean "being only one out of many" for Copernicus originally, as far as I know. Setting these "definitions" equal is an extension of the original principle introduced later by others.

I don't accept the only-one-out-of-many version of the Copernican principle in certain cases, especially when it comes to extra-terrestrian and simulated intelligent people.

I wouldn't be surprised, if there are indeed civilizations built by intelligent life on planets around some stars. But I would be surprised if these civilizations were not very different in many aspects of higher behaviour, society, culture -- yes: civilization. As we can see here on earth already, many of these aspects are coupled only very losely or not at all to the physical and biological basics.

The "civilizational" variation among extra-terrestrians will be immense, I think, and because of this each of these civilizations -- including our own -- will be something particular.

What's about simulated intelligent people? If we live in a simulation run by our descendants, then these descendants will most likely live in a civilization very different from their ancestors. Because of this, we know nearly nothing or nothing at all about the reasons for running this simulation. The simulated people may be substantially different from the real ancestors of the people running the simulation -- the simulated people are something particular. And if somebody wants, "substantially different" may include "not occuring in great numbers".

This all may be seen as an unfounded speculation too, of course ;-)

Martin Andersen said...

Indeed, it's all pure speculation :-)

We can't say much about those who simulate us, other than they must have technology way beyond what we currently have.


I don't think the simulation scenario could solve the Fermi paradox, as they may or may not simulate aliens..

Ranman said...

I really don't think we'll ever see a fully simulated universe because to simulate everything you would need just as much energy as the simulation took... so every time you simulated it, the compression, the lossy universe would get smaller and less realistic each time until eventually you wouldn't have enough power to simulate another universe.

So unless there is some undiscovered law of physics (and trust me it's unlikely at this point) that creates limitless space and energy then you'll never be able to simulate everything... not down to the quark level... not even down to the electron level... so at best you would have really shitty collision detection (not noticeable to us but obvious under experimentation)...