Friday, November 15, 2013

Skepticism, Godzilla, and the Artificial Computerized Many-Branching You

Nick Bostrom has argued that we might be sims. A technologically advanced society might use hugely powerful computers, he says, to run "ancestor simulations" containing actually conscious people who think they are living, say, on Earth in the early 21st century but who in fact live entirely inside an advanced computational system. David Chalmers has considered a similar possibility in his well-known commentary on the movie The Matrix.

Neither Bostrom nor Chalmers is inclined to draw skeptical conclusions from this possibility. If we are living in a giant sim, they suggest, that sim is simply our reality: All the people we know still exist (they're sims just like us) and the objects we interact with still exist (fundamentally constructed from computational resources, but still predictable, manipulable, interactive with other such objects, and experienced by us in all their sensory glory). However, it seems quite possible to me that if we are living in a sim, it might well be a small sim -- one run by a child, say, for entertainment. We might live for three hours' time on a game clock, existing mainly as citizens who will give entertaining reactions when, to their surprise, Godzilla tromps through. Or it might be just me and my computer and my room, in an hour-long sim run by a scientist interested in human cognition about philosophical problems.

Bostrom has responded that to really evaluate the case we need a better sense of what are more likely vs. less likely simulation scenarios. One large-sim-friendly thought is this: Maybe the most efficient way to create simulated people is to evolve up a large scale society over a long period of (sim-clock) time. Another is this: Maybe we should expect a technologically advanced society capable of running sims to have enforceable ethical standards against running small sims that contain actually conscious people.

However, I don't see compelling reason to accept such (relatively) comfortable thoughts. Consider the possibility I will call the Many-Branching Sim.

Suppose it turns out the best way to create actually conscious simulated people is to run a whole simulated universe forward billions of years (sim-years on the simulation clock) from a Big Bang, or millions of years on an Earth plus stars, or thousands of years from the formation of human agriculture -- a large-sim scenario. And suppose that some group of researchers actually does this. Consider, now, a second group of researchers who also want to host a society of simulated people. It seems they have a choice: Either they could run a new sim from the ground up, starting at the beginning and clocking forward, or they could take a snapshot of one stage of the first group's sim and make a copy. Which would be more efficient? It's not clear: It depends on how easy it is to take and store a snapshot and implement it on another device. But on the face of it, I don't see why we ought to suppose that copying would take more time or more computational resources than evolving a sim up from ground.

Consider the 21st century game Sim City. If you want a bustling metropolis, you can either grow one from scratch or you can use one of the many copies created by the programmers or users. Or you could grow one from scratch and then save stages of it on your computer, shutting the thing down when things don't go the way you like and starting again from a save point; or you could make copied variants of the same city that grow in different directions.

The Many-Branching Sim scenario is the possibility that there is a root sim that is large and stable, starting from some point in the deep past, and then this root sim was copied into one or more branch sims that start from a save point. If there are many branch sims, it might be that I am in one of them, rather than in a root sim or a non-branching sim. Maybe one company made the root sim for Earth, took a snapshot in November 2013 on the sim clock, then sold thousands or millions of copies to researchers and computer gamers who now run short-term branch sims for whatever purposes they might have. In such a scenario, the future of the branch sim in which I am living might be rather short -- a few minutes or hours or years. The past might be conceptualized either as short or as long, depending on whether the past in the root sim counts as "this world's" past.

Issues of personal identity arise. If the snapshot of the root sim was taken at root sim clock time November 1, 2013, then the root sim contains an "Eric Schwitzgebel" who was 45 years old at the time. The branch sims would also contain many other "Eric Schwitzgebels" developing forward from that point, of which I would be one. How should I think of my relationship to those other Erics? Should I take comfort in the fact that some of them will continue on to full and interesting lives (perhaps of very different sorts) even if most of them, including probably this particular instantiation of me, now in a hotel in New York City, will soon be stopped and deleted? Or to the extent I am interested in my own future rather than merely the future of people similar to me, should I be concerned primarily about what is happening in this particular branch sim? As Godzilla steps down on me, shall I try to take comfort in the possibility that the kid running the show will delete this copy of the sim after he has enjoyed viewing the rampage, then restart from a save point with New York intact? Or would deleting this branch be the destruction of my whole world?

29 comments:

Marco Devillers said...

To be a Non-Player Character, or not to be!

Anonymous said...

I would treat identical copies or branches of me with the same consideration as I treat any of my future versions that arise from the natural progression of my life. At least I think that's the most sensible way to look at it. I would even treat others as "copies of me with variation": People are identical to the degree to which they share properties.

Of course, this is not how practical social identity works in society, it is merely a philosophical consideration, however one with ethical implications. I would pretty strongly care about the welfare of my copies, and seeing other beings partly as my copies grounds utilitarian intuitions. I would hope that a civilization capable of running conscious simulations shares this sentiment sufficiently to motivate at least minimal ethical standards for them.

Eric Schwitzgebel said...

Marco: LOL.

Anon 09:10: That's my initial inclination too, but if you look at the personal identity literature in philosophy and their treatment of branching teleporter cases, it's pretty unorthodox. And in a more "traditional" branching case (I step into the teleporter on Earth and then my Earthly body is destroyed and duplicates appear simultaneously on Mars and Venus), I definitely feel the pull of the idea that my Venus-self would not think of the Mars-self as "me".

Anonymous said...

I hope (perhaps too optimistically) that both my Mars and Venus versions would consider each other "me", enabling cooperation in the prisoner's dilemma without external pressure. On the other hand, it's easy to see this breaking in the case of real conflicts, maybe fueled by akrasia or tragic misunderstandings.

I'm surprised it's the unorthodox view. My problem with the alternative views is that I just don't see that much coherence within one lifetime either. The intrapersonal, intertemporal conflicts are easily visible in the existence of regret, time preferences, failing commitment devices and changing values. On the level of consciousness, it's a trivial insight that experiences are local in time, and memories and predictions are only mediocre forms of temporal integration.

Add hypothetical time travel, and the divergence in time is just as obvious as the divergence of branching copies. (The movie "Time Crimes" is an excellent exploration of how personal agency can tragically fold in on itself by accidental time travel). At the end of the day, the future is just another place in spacetime.

Carl M. said...

Here's an idea I've been playing with: "To be is to cause. To experience is to be caused."

Under this idea, how can we tell what's real from what's not? Simple, what is real is causally efficacious, and what's not is not.

A unicorn for example never causes unicorn poop, because unicorns aren't real, but the idea of unicorns has caused many Lisa Frank trapperkeepers because the idea of unicorns is real.

So, in a sim world, what's real? It depends on how the sim is programmed.

If the sim is lazily programmed, it may look like A causes B, but the actual causes are quite different. For example, it may look like the wall in a video game causes the character to stop, but in reality, characters are caused to stop by their collision detection, and we often see videogames where hands, swords, guns, hair, etc. pass through other objects due to poor collision detection.

In this world, I think that my actions are caused by "me" and that "I" am real, but if it turns out that my actions are entirely caused by some internal calculations of the simulation without respect to what we think of as ordinary causal patterns, then it turns out that I am not real.

Causation, of course, always overlaps, but I think this is a good way of telling what is or isn't real.

Jen said...

Hi Eric,

I'm a science person, but I've toyed with philosophy a bit in a number of courses. I've encountered the Matrix idea before as well, and I see its skeptical validity, including in applying a simulation label to life. I think the simulation argument adds an interesting dimension to all-possible-worlds thinking. It reminds me a bit of a choose your own adventure novel, where there are several stories within the same book. I imagine this is somewhat what branches of self must be like - somehow both distinct.

However, what I'd like to do is propose an alternative, instead of the possibilities within the sim.

I considered the Matrix possibility once, having a peculiar experience which led me to see that although I did not always grasp it, there was a mathematical computation underlying the universe. I was in junior high, and staring at the frost on the pane of a school bus, and I starting just computing numbers in my head, from staring at the frost. What I "saw" was a mathematical function of a pattern of differences in squares of numbers. (I asked a statistics teacher at some point in the future if someone had noticed this.) I could look at the fact that all of these symetrical forms, and numerically computational entities exist in the universe, frost, snowflakes, genes, possibly quantum patterns of thought, and consider the possibility. Computers are able to do computations in random sequences, but all the random sequences have an eventual pattern. And, the universe? If people could do it, then it's quite possible that we are like the egg within the egg concept - nested thinkers, with larger thinkers outside of ourselves. I think the best thing this says, is that humans are not so egoistic as they once were. In terms of the possibility, I could just as easily declare myself the goddess, go around declaring free will and taking my power back, much like the empowered Matrix character. Depending on the level I take this to, I could also be declared deluded, and hospitalized.

The problem with all of these concepts: Matrix, simulation, evil scientist, controlled by gods, etc., is that the theory of skepticism that philosophers consider in which we give up control or power, goes along with the technology of the times. If in reality, we were controlled by a futuristic society, earlier generations would have had an indication of some technological possibility before we had created computer simulations as a social advancement. Rather, the considerations of these types of things advanced with our technology.

I think the larger possibility, is that when understanding the mathematical and pattern nature of the world, some genius said "I want to create a machine that does what I see." Einstein was thought to have intuited something when he decided to test the Theory of Relativity. If we look at a larger pattern, perhaps what we're intuiting is really just us - people, as a whole. We start down at a pattern level within cells, and then interactions of cells that form a structure, structures that form systems, systems that form a person. Maybe beyond that, we intuit a connection between people and other entities in the universe that forms a pattern of one large whole living entity. So, to be aware of the pattern is not somehow supernatural, alien based, but rather to pick up on a being that a person is a sytem within. Not something outside of ourselves, controlling existence. We have the control from inside, just as I cannot do anything if something happens to my hippocampus on the inside. It's going to affect my memory. I may have some outer control, based on top down processing, depending on the extent of the injury. So, from the inside, it's bottom up, with occasional intuiting from other systems within the space of existence.

Eric Schwitzgebel said...

Thanks for the continuing comments, folks! I'm rushing around New York and then catching a flight home (unless Godzilla gets me first) so it might be a couple days before I have a chance to think them through and reply....

Callan S. said...

All the world's a stage?

clasqm said...

"simulation" is a loaded term. Creating a simulation of their own universe might be ONE thing these alleged superbeings might be interested in doing, but why should dredging up their own past be their overriding interest when there is an infinitude of possible realities to invent? I mean, we do make movies that are not historical dramas, don't we? Why shouldn't they?

In other words, our relation to the superbeings mght not be that of a Sims character to a game player. We might be no more than a Tetris block.

(the captchas are completely unreadable today - does that mean the program is coming apart?)

Eric Schwitzgebel said...

Anon Nov 15: I definitely feel some of the attractions in that line of thought. On the other side, consider this: Suppose Mars-you commits a crime two years after divergence from Venus-you. He has settled in with some shady characters that have changed him for the worse. Venus-you, meanwhile, has remained an upstanding citizen. How would you feel about the police arresting Venus-you rather than Mars-you (who is now on the lam and hard to find), arguing that after all it was YOU who committed the crime! -- as much a part of you as a 5-year-old Earth you was.

Eric Schwitzgebel said...

Carl M: Some metaphysicians would accept a principle somewhat like yours, e.g., some interpretations of Nietzsche. I'm not sure that the computer program that constitutes sim-me is any less "me-ish" as a cause than is the flesh and brain that make up non-sim me. But it does seem that the sim possibility makes it easier to conceive of causal streams not really involving me directly and that hijack my behavior, like an outside program that tweaks my preferences and decisions. How to delineate what is inside vs. outside becomes a little problematic, though....

Eric Schwitzgebel said...

Jennifer: That reality has mathematical patterns at root is a pretty attractive idea! I would take issue with one of your statements though:
"If in reality, we were controlled by a futuristic society, earlier generations would have had an indication of some technological possibility before we had created computer simulations as a social advancement."
I don't see why this would have to be the case.

Eric Schwitzgebel said...

Callan: In a sense, yes!

Michel: Right. I see no reason to think they would make *only* ancestor sims. The question is whether they make enough of them that the population of simulated people approaches or exceeds the population of non-sim people.

Captchas coming apart seems to me only very weak evidence that the universe is coming apart. Some alternative explanations seem more epistemically attractive to me! (Does that mean it's *no* evidence? I'm not sure of that.)

Anonymous said...

How would you feel about the police arresting Venus-you rather than Mars-you
One function of imprisoning criminals is to prevent them from committing further crimes during their imprisonment. This function would fall away in this case, unless there is estimated to be a high probability that Venus-me will undergo a similar criminal career.

As for the deterrence function, it depends on whether Mars-me sees himself as sharing an identity with Venus-me. If so, the deterrence should work equivalently to the normal deterrence of criminals who empathize with their normal future selves. In dealing with people who hold the (majority) view that they don't share an identity, deterrence would be reduced and the punishment probably perceived as unjust.

I can intuit the problem by regarding family members. If I had to pay off my father's debt or go to prison for my brother's crime, I would have a problem with that. But this is largely because the normal genetic variation between people is considerable even within families - their choices don't reflect my personality, and so I share less of an identity with them.

On the other hand, if it could be proven to me that time travel is real and in an alternative timeline or in the future, I will commit a crime or cause a cost that I could see myself committing or causing in the respective context, treating me as responsible seems much more fair and sensible to me. After all, I would also be held responsible for my actual past selves' debts and choices, and there's a meaningful way to argue I'm no longer them.

Jen said...

Eric, regarding your comment: So, you believe that in the creation of conscious beings by a futuristic society, everyone would have consciousness, but not one of them would have an awareness in the peripheral sense of an outer force? I could believe most people would have streamlined and focused attention on the immediate surroundings of the simulation, but I find it hard to believe that not one being would pick up on the source of the creation, and the technological reality.

Juan said...

How does the multiverse figure in this? If we were living in one I would assume that the probability of this scenario would increase exponentially. Am I wrong in thinking that? What does Bostrom say about this (if anything)?

Also, taking this further, I would argue that there's a non-trivial chance that any kind of possible fiction has been realized in a computer and is effectively real to the people living inside it. Realistic scenarios wouldn't be only ones to be made; as you say, there's always a possibility that Godzilla will come bursting through your window. So I foresee branching sims as places where people would play out fantasies in addition to the "normal" sim experiments.

Callan S. said...

Eric, any chance of seguing this over to Scott Bakkers blind brain hypothesis - some kind of virtual world the shape of which sort of perfectly matches the scope and depth of the processor which takes in that virtual world.

Callan S. said...

Jennifer, possibly if the simulation used an evolutionary model, it would remain utterly consistant. Because whatever evolved in it would be only the things that survived - they would not rely on any notions or perceptions which are actually a reflection artifact of a VR creators notions and perception.

But maybe even in such a situation, if one looks back far enough, one might find the alien to the VR universe notions and perceptions showing up. Ideas that don't work here but do work in the universe the VR creator is in. But they would be hidden under billions of years of VR world consistant behaviour.

Go back far enough and maybe you find miracles?

Jen said...

Interesting concept Callan. Thanks!

Anonymous said...

"Realistic scenarios wouldn't be only ones to be made; as you say, there's always a possibility that Godzilla will come bursting through your window."

Reminds me of the apartment demolition scene in Stranger Than Fiction.

Eric Schwitzgebel said...

Thanks for the comments, folks!

Jennifer/Callan: Right, the main thing is that there be (as Juan puts it) a "non-trivial chance" of the people in our sim being ignorant of their status. That's consistent with evidence of simulation in the majority of sims.

Anon Nov 19: It's quite a tangle of issues, as you're nicely laying out! Mere similarity doesn't seem enough to justify punishment, though; you presumably want *identity* with the perpetrator (and probably also the crime in the past), but straightening that out is pretty tricky!

Juan: The multiverse possibility dramatically increases the likelihood of a sim somewhere, but it's not clear what it does to the probability calculation, since infinitude is so weird.

Callan: Yes, I do think this can be hooked up nicely with Bakker -- probably both with his brain-blind hypothesis and with his fictions.

Anon Nov 20: Thanks for the tip on Stranger than Fiction -- I haven't yet seen it.

Juan said...

I've been thinking if there is some sort of way sims could works out they are in a sim. You talk about ancestor sims; what happens if the historical reconstruction isn't entirely accurate? Information about the past gets lost all the time. What if the scientists running the experiment took liberties with the sim that would be obvious to a real contemporary?
It also stands to reason that the further back you go, the less information you have to go on and the less accurate the sim will become. But the sims living in those times would find it more difficult to work it all out due to their development, and prob rationalize glitches as "magic" or some such. But to more intelligent beings maybe there is some (mathemetical?) by which they could find all the gaps.

Eric Schwitzgebel said...

Juan, right, you could take gaps or inconsistencies of a certain sort as evidence that one is in a sim -- but which gaps an inconsistencies? Would the weirdness of quantum mechanics count? And the *lack* of gaps and inconsistencies is consistent with its just being a well-executed sim or a certain type....

Juan said...

You're of course right in pointing out that there is a lot of ambiguity in deciding what counts as a "true" inconsistency, which is what I hoped to get at. I think unintuitive "weirdness" is impossible to judge, because as odd as a physical fact may seem, like QM, it's still logically consistent. Ditto for unusual historical events (the Kennedy assassination befuddles everyone, but it's not beyond the bounds of possibility).

Another thought: since sims are essentially programs run on computers, aren't they always at the risk of bugs? No matter how "well-executed" a sim may be, if it's like any kind of computer that we know of it's bound to encounter coding errors at some point. Human level intelligences wouldn't clearly be up to the task of running large scale sims for long periods of time; I think we have to assume super intelligent beings are behind any well constructed and enduring sim. But error still seems like a real possibility, it may just take a while to manifest itself. The point of all this being is that inconsistencies can appear regardless of how brilliantly designed the sim is, although how often and big that anomaly can be is dependent on the IQ of the designers.

Eric Schwitzgebel said...

"Bugs" is an interesting issue. I see several way to deal with that issue from within the sim possibility: (1.) As you suggest, super-intelligent beings could have designed a bugless sim. (2.) Bugs could appear but they could manifest as miracles, insanity, unexplained phenomena, or paranormal phenomena. (3.) Bugs could appear but we could be designed so as not to notice them or so as to immediately forget them. (4.) We might be living in a sim with a very short past and our impressions of the past might be designed so as to seem bugless. (5.) We might be living in a very well-tested sim.

Eric Steinhart said...

I'm late to this party, but it doesn't seem like all the computer stuff adds anything besides high-tech decoration to the underlying metaphysics or logic. It's just branching time / branching worlds. So the other Erics are your counterparts.

Eric Schwitzgebel said...

Thanks for the comment, Eric! I agree that it's a branching-worlds case, but I don't think the high-tech stuff is just decoration. I think it influences our sense of what is possible and it influences our intuitive judgments. So I have a pretty firm intuition that my counterparts in Everett branching worlds are not me, and I wouldn't sacrifice my own interests for their benefit. My intuition is much less clear in simulation cases -- I'm not entirely sure why, yet, though!

Eric Steinhart said...

Of course your counterparts aren't you. By definition, they aren't you. Nor is the you of five minutes ago the you of right now.

As for concern, who cares? People persisting in perfectly ordinary natural biological ways fail entirely to show concern for their future selves (i don't smoke that much; one more drink can't hurt; no seat belt for me). Concern about the future, like stock market investing, does little more than reveal our irrationalities. It doesn't track anything deeper.

Eric Schwitzgebel said...

Interesting perspective, Eric! The view is appealing for its metaphysical cleanliness, though distasteful in its implications. I've taken the liberty of emailing you a very short story I've written in which the central character adopts a philosophical perspective something like the one you express here.