Pseudo-Criticism

Years of studying the history, philosophy, and sociology of science led me to conclude that the word "science" says little useful beyond "good research."  Peter Woit over at Not Even Wrong demonstrates, ranting against Bostrom’s "pseudoscience" simulation argument.   His complaints: he doesn’t see how to check it with data soon, it is easier than his research but has "dense thickets" of reasoning he finds too hard to follow, and it is not very connected to and distracts attention from his research areas: 

On the pseudo-science front … beyond the edge of absurd, there’s today’s NYT Science Times section, which features a piece by John Tierney about the ideas of philosopher of science Nick Bostrom. … that there’s a significant probability that our universe is just a simulation being conducted by a more advanced civilization … Maybe we should be trying to entertain our creators so they will not turn off the simulation? …

The main reason I find myself getting annoyed with discussions of it here is that generally it’s pretty irrelevant to the science I’m concerned about.  Not only that, but a huge amount of damage is being done to that science by an increasingly large number of people who seem unable to tell the difference between science and science fiction. … people who want to do pseudo-science because it’s a lot easier than science will keep on justifying absurd, and inherently untestable speculation, claiming that "how do you know that a miracle won’t happen if we work on this? If we do, maybe we’ll find a real test!"

People who do this behave exactly the same way as every crackpot I’ve ever made the mistake of arguing with, trying to draw you into an endless investigation of the dense thickets of their idiocy. Arguing with someone who thinks the "simulation argument" is a scientific hypothesis is just this kind of waste of time.

… I don’t see what the problem is with "lumping Bostrom’s ideas in with religion".  They’re not science and have similar characteristics: grandiose speculation about the nature of the universe which some people enjoy discussing for one reason or another, but that is inherently untestable, and completely divorced from the actual very interesting things that we have learned about the universe through the scientific method. …

The thing which is likely to lead me sooner or later to have to give up and shut down this blog is … the large number of people who want to turn this into a discussion forum for crackpottery and various forms of pseudo-science.

Sure there are possible ways you could define "science"; but few agree on which one, and when a proposed definition seems to conflict with "good research" people usually start to look for another definition.

Added: Woit responds with "neener neener." 

GD Star Rating
loading...
Tagged as:
Trackback URL:
  • http://dmorr.livejournal.com/76003.html Dave Orr

    Woit’s complaints seem strangely off base. His main issue is that he doesn’t think Bostom’s ideas are science? Boston is a philosopher! His topic is philosophy of science, and at no point did he claim that he was doing science as part of his simulation thought experiment.

    That’s like complaining you don’t like penguins because they’re not ice cream.

    • http://www.martinpeniak.com Martin Peniak

      well said, completely agree, I did not find the criticism very useful but rather full of biases…

  • http://pdf23ds.net pdf23ds

    Yeah, what Orr said. Not for a second did I ever think Bostrom was attempting science with his simulation argument. It’s philosophy, through and through. We might as well claim Darwin was practicing pseudo-philosophy in The Origin of Species.

  • Constant

    Dave Orr and pdf23ds – it’s in the Science section of the New York Times. Putting there implicitly labels it as science, so it is legitimate to treat it as something that is being passed off as science, and to quarrel with that.

    Whether Woit has managed to come up with a good argument is a separate matter; Robin Hanson seems to think not.

  • http://www.michaelparente.net Michael

    So, now we have to define ‘good research’. I’d say it includes testable hypotheses and experimental evidence. Everything else is a memoir.

  • http://profile.typekey.com/robinhanson/ Robin Hanson

    Folks, “philosophy” is no clearer a descriptor as “science.” Rather than complain that something is not “science,” or not “philosophy,” it is much better to just say more specifically what it is that you don’t like about it.

  • anonymous

    Whatever it is, it’s definitely not pseudoscience or religion, as Woit has characterized it. He shows no signs of retracting his comments, or even responding to comments about this over at his blog (I notice that Robin Hanson’s comments over there have been ignored).

  • michael vassar

    Woit, like many people critical of string theory, are probably not very interested in epistemology or metaphysics and have a knee-jerk negative reaction to the word “anthropic” due to it’s association with string theory and the abuses that they attribute to the string theory community.

  • http://pixnaps.blogspot.com/2007/08/scientism.html Philosophy, et cetera

    Scientism

    Many otherwise-intelligent people fail to appreciate the possibility of non-scientific rational inquiry, i.e. philosophy. The prevalence of ignorant scientism in this thread (bashing Nick Bostrom’s simulation argument) is remarkable — though sadly not…

  • http://pixnaps.blogspot.com Richard

    Hmm, my trackback doesn’t seem to be working, but anyway… I offer some further criticisms here.

  • http://magicdragon.com Jonathan Vos Post

    Nor is Dr. Woit interested in the prior literature on the subject, which discredits Bostrom’s claims to precedence. Woit declined to post the following submitted comment.

    The Establishment approach to eliminating crackpots is to restrict to
    authors of refereed papers, and further, in the disciplines in which
    they are published (to eliminate the phenomenon of an expert in one
    field pontificating spuriously in another field).

    That eliminates a small number of famous rejected but significant
    papers, and lets in a few goofy things that referees mysteriously
    okayed, but is generally useful to, say, academic bloggers.

    The background of my article “Human Destiny and the End of Time”, in
    the magazine Quantum (edited by a PhD Chemist, by the way) and
    published years before Bostrom’s rediscovery, was my own prior
    publications on extrapolating the advance in computing power.

    Jonathan V. Post, “Quintillabit: Parameters of a Hyperlarge Database”,
    Proceedings of the Sixth International Conference on Very Large
    Databases,
    Montreal, Canada, 1-3 October 1980

    The above coined the term “1 Shannon = 1 mole, Avogadro’s number, of bits.”

    Simulating a person, or the entire universe, is estimated in Shannons.
    How many Shannons do you think it takes to simulate the obervable
    universe to, say, femtometer and femtosecond resolution?

    My PhD work included analyses of how many bytes per second are
    processed in a single cell (bacterium or yeast) and extrapolating to
    how many computations per second are needed to simulate an entire
    organism in ultra-high resolution. The chapters of that dissertation,
    arguably the first ever on what became Nanotechnology, were read
    worldwide when repackaged as refereed papers, and expecially intrigued
    scientists at U. Moscow and U. Leningrad, which the State Department
    preferred I not correspond (this being the Cold War, and my doing
    highly classified work for Boeing, Rockwell, Lockheed, Army, Navy, and
    Air Force).

    The government explanation to me was: “We consider your research to be
    bullsh*t. But on the off chance that it is not, we don’t want you even
    to send repreints of existing literature to the communists.” Then they
    asked if they could have the Bulgarian postage stamps from some
    letters they intercepted, inviting me to chair a session at an
    international conference in Sofia, on the basis that one of these Men
    in Black had a son who collected stamps. I refused, rather curtly.

    In any case, I am not asserting that I published PART of what made
    Bostrom famous. I’m asserting, with complete support from edited
    literature, that I discovered ALL of what he later claimed, and that I
    left out the more foolish parts of his quasi-religious hand-waving.

  • http://julesandjames.blogspot.com/ James Annan

    Robin,

    In both of your comments at NEW, you talk about “checking” the hypothesis, and seem to think that Peter’s complaint is that it might take a long time to “check”. Can you propose any method by which it can be “checked”?

  • http://magicdragon.com Jonathan Vos Post

    Mind you, I enjoyed Dr. Woit’s book, and read his blog daily. But if we accept, for the sake of argument, his claim about “damage is being done to that science by an increasingly large number of people who seem unable to tell the difference between science and science fiction” — then it would seem important to consider the opinions of those people who are experts in Science Fiction AND in Science. They (or, more properly, we) are better at telling the difference. Would Woit delete comments from Dr. Isaac Asimov (if Asimov were still alive? Would he delete a comment from Sir Arthur C. Clarke? Would he delete a comment from Dr. Gregory benford? Well, he did delete the following comment submission.

    Not Even Wrong
    More of the Usual Sorts of Things
    http://www.math.columbia.edu/~woit/wordpress/?p=583#comment-27459

    [truncated]

    # Peter Woit Says:
    August 15th, 2007 at 12:21 pm

    I’ve deleted several comments of the “SF sucks!” “No it doesn’t!”
    variety, and don’t want to really encourage general discussion of SF
    here, something I have very limited interest in.

    still waiting…

    Sorry to report that it doesn’t look like I’ll be writing up anything
    at an undergrad level about math and QM any time soon. This year I’ll
    be teaching our representation theory graduate course, and
    concentrating on working on that and on research. Maybe sometime in
    the future I’ll get involved in trying to teach this topic to
    undergrads, then would write something up.

    # Garbage Says:
    August 15th, 2007 at 12:43 pm

    “Anyone who thinks it is a good idea to discuss these questions
    seriously is encouraged to do so at Tierney’s site, not here.”

    I wonder Peter why do you mention it in the first place? ;)

    Sad to hear Julius Wess is gone, he may have taken SUSY with him, we
    will soon find out…

    # Jonathan Vos Post Says: Your comment is awaiting moderation.
    August 15th, 2007 at 11:07 pm

    With all due respect, Dr. Woit, it is NOT a matter of “SF sucks!” “No
    it doesn’t!”

    I understand why you don’t want to really encourage general discussion
    of SF here. But your admission that Science Fiction is “something I
    have very limited interest in” is precisely why you could learn from
    commenters such as John Baez who are VERY good at mathematcs and
    Physics, and know how Science Fiction writers (some of whom are
    professional mathematicians and Physicists) have speculated on some of
    your concerns.

    The Math and Physics community is poorer when paying no attention to
    the wider Literature that often discussed deep issues decades before
    the formal literature caught up.

  • http://quantumfieldtheory.org nc

    I’ve been programming simulations since 1982. It’s impossible to simulate with complete accuracy any multi-body situation. The whole of mathematical physics is contrived to either deal inaccurately with multibody situations (e.g., statistical models in quantum mechanics) or to only deal with two bodies.

    I remember being shocked when I learned quantum mechanics at college to find that you can only get analytic statistical solutions for the hydrogen atom, and for helium and everything else the statistical solutions are approximate. In other words, there’s no determinism at all even for hydrogen, and for heavier atoms the situation is that you can’t even write down an exact mathematical non-deterministic model, instead you have to make do with various approximations.

    The relevance of this to “computer simulations of the universe” is that there are 10^80 atoms in the universe, and you can’t get exact solutions for even a single atom in practice! Then think about simulating nuclear physics and the fact that QCD is so complex with gluon and quark screening and anti-screening effects. Even if the modelling of a single atom by computer requires only a computer of 10^25 atoms, to simulate all the atoms in the universe would take a computer containing (10^25)*(10^80) = 10^125 atoms. This simulation idea is extremely extravagant: the whole idea of a computer simulation in the real world is to save the time and money required to do the real experiment. It would take far less effort and energy to create the universe than to build a computer capable of simulating the universe. Hence the hypothesis full of problems. Mathematics is over-hyped and people religiously believe the universe is represented by mathematical models that can be solved by computer, which is not true:

    ‘… the ‘inexorable laws of physics’ … were never really there … Newton could not predict the behaviour of three balls … In retrospect we can see that the determinism of pre-quantum physics kept itself from ideological bankruptcy only by keeping the three balls of the pawnbroker apart.’

    – Dr Tim Poston and Dr Ian Stewart, ‘Rubber Sheet Physics’ (science article) in Analog: Science Fiction/Science Fact, Vol. C1, No. 129, November 1981.

    ‘… electrons: when seen on a large scale, they travel like particles, on definite paths. But on a small scale, such as inside an atom, the space is so small that there is no main path, no ‘orbit’; there are all sorts of ways the electron could go, each with an amplitude. The phenomenon of interference [due to pair-production of virtual fermions in the very strong electric fields (above the 1.3*10^18 v/m Schwinger threshold electric field strength for pair-production) on small distance scales] becomes very important, and we have to sum the arrows to predict where an electron is likely to be.’

    – R. P. Feynman, QED, Penguin, 1990, page 84-5.

    ‘It always bothers me that, according to the laws as we understand them today, it takes a computing machine an infinite number of logical operations to figure out what goes on in no matter how tiny a region of space, and no matter how tiny a region of time. How can all that be going on in that tiny space? Why should it take an infinite amount of logic to figure out what one tiny piece of spacetime is going to do? So I have often made the hypothesis that ultimately physics will not require a mathematical statement, that in the end the machinery will be revealed, and the laws will turn out to be simple, like the chequer board with all its apparent complexities.’

    – R. P. Feynman, The Character of Physical Law, November 1964 Cornell Lectures, broadcast and published in 1965 by BBC, pp. 57-8.

    Regarding Mr Woit’s “rants” or rather “facts”, you know, religion (in the wider sense) is a falsifiable science: merely commit suicide and see if you go to heaven or to hell… (maybe you would do us a favor by shuting up and actually try that nice little scientific experiment, I’ll promise to nominate you for a Nobel if you discover something!).

    String theory contains 6 unobservably small dimensions whose exact sizes and shapes determine the quantitative parameters of particle physics if string theory is right. There’s a landscape of 10^500 different sets of predictions, and nobody has identified even one of those that is even an ad hoc model of what is known already.

    Similarly, if you have a (fiddled) ‘theory’ which comes in two forms, one of which ‘predicts’ a coin will land heads and the other ‘predicts’ tails, then we toss the coin, and the theory is right whatever happens. This is not science, and trying to obfuscate the fact it is a lie to call it science, by referring to the anthropic principle, is quackery. The top dogs in string make money from selling pseudo-physics.

    The fact is that it will not merely take a ‘long time’ to find out if there are 6 rolled up extra dimensions in a Calabi-Yau manifold in each particle. It will never happen, even if you could probe the Planck scale, the uncertainty principle tells you that you will be disturbing what you are probing and you won’t be able to accurately find out the state of the Calabi-Yau manifold and the precise way the extra dimensions are compactified. It’s a delusion. You’d need a particle accelerator the size of the universe to try, anyway.

    There is a lot of data that need to be modelled in particle physics, and string theorists do nothing about it:

    ‘… I do feel strongly that this is nonsense! … I think all this superstring stuff is crazy and is in the wrong direction. … I don’t like it that they’re not calculating anything. I don’t like that they don’t check their ideas. I don’t like that for anything that disagrees with an experiment, they cook up an explanation … All these numbers [particle masses, etc.] … have no explanations in these string theories – absolutely none!’ – Richard P. Feynman, in Davies & Brown, ‘Superstrings’ 1988, at pages 194-195, http://quantumfieldtheory.org/

    They are liars, charlatans, quacks and evil shits.

  • http://evolutionarydesign.blogspot.com island

    Woit, like many people critical of string theory, are probably not very interested in epistemology or metaphysics and have a knee-jerk negative reaction to the word “anthropic” due to it’s association with string theory and the abuses that they attribute to the string theory community.

    We have a winner, folks. Not that I support Bostrom’s position, but Peter has been acting in a purely reactionary manner toward the AP ever since Danny Lunsford showed back up on the scene.

    No surprise there..

    I am the author of http://www.anthropic-principle.ORG and http://evolutionarydesign.blogspot.com

    Check em out.

  • Nick Tarleton

    nc, a computer simulation of the universe wouldn’t have to model everything on a quantum scale. It would just have to achieve sufficient accuracy that we embedded observers couldn’t tell the difference, so it could get away with, e.g., modeling most of Earth using Newtonian physics (and not even bothering with atoms).

  • anonymous

    “They are liars, charlatans, quacks and evil shits.”

    Really? I saw Lisa Randall on TV today, and she seemed like a really nice person, not an evil shit. Perhaps you are an evil shit?

  • anonymous

    http://www.math.columbia.edu/~woit/wordpress/?p=586

    Woit has issued a sarcastic response to these comments.

  • Nick Tarleton

    More substitution of rhetoric for thought, eh? Shame.

  • http://raghavgupta.wordpress.com Fractal Folds

    Ok, lets say the WE-ALIAS (WE Are Living In A Simulation) theory is true.

    Now what?

  • Anonymous

    Hey, maybe we’re actually part of a multi-player simulation in which we’re not controlled by one super-nerd running a simulation on a super-computer, but by many super-nerds competing in a networked-game on many super-computers. And imagine further, that those numerous super-nerdy-super-computer simulations are actually a simulation run by a bunch of other even nerdier super-nerdy-super-computer geeks running parallel simulations.

    Oh, hold on for a minute, the whole idea is completely retarded. You guys are a bunch of intellectual hacks. Go smoke some your weed and leave the intellectual work to intellectuals. You guys give higher education, and especially philosophy, a bad name.

  • http://profile.typekey.com/robinhanson/ Robin Hanson

    Hey newcomers, we don’t like such long comments here. Feel free to post a longer essay elsewhere and link to it from the comments. Jonathan and nc, you could have made your basic point in a few paragraphs.

    And everyone, watch your language.

    James, with time we will learn just how feasible and popular are the postulated simulations.

  • http://www.stafforini.com Pablo Stafforini

    Oh, hold on for a minute, the whole idea is completely retarded. You guys are a bunch of intellectual hacks. Go smoke some your weed and leave the intellectual work to intellectuals. You guys give higher education, and especially philosophy, a bad name.

    Bostrom’s paper has been published in a top philosophical journal and has been discussed by academics working in top philosophy departments. If you are going to dismiss the simulation argument on the grounds that it is bad philosophy, you need to explain the fact that the expert philosophical community doesn’t think so. Merely dismissing “the whole idea” as “completely retarded” simply won’t do.

  • http://julesandjames.blogspot.com/ James Annan

    Robin,

    I think you miss Peter’s point (and many others’, I suspect) that there is no evidence that could ever serve to show that we are _not_ living in a simulation.

    Now I see that the simulation argument is slightly different from belief in god, in that the argument involves our own future ability to create a simulation, and the argument is not actually a belief in simulations, but a logical relationship between 3 statements. But do you actually think there is any way we could show that such a simulation is not possible? What evidence would indicate this? Can we not always argue that the improved technology is just another few decades away?

    How about the following “simulation argument”: Given time, humans and all other advanced societies may develop supernatural powers of creation, and become gods. Either (1) this will not happen (2) we decide not to use these powers or (3) we are likely living in a world created by (a) god.

    Is this an interesting scientific theory worthy of discussion in the NYT? Maybe I should write it up in longer form :-)

    I’m not arguing that this is bad philosophy or should not be said. I’ve got a lot of sympathy for those who say that such fundamentally unfalsifiable theories are not scientific though.

    The counterargument that we might actually create a simulation, thus directly demonstrating its feasibility, is of course not relevant. We might also become supernatural, or some supernatural being may appear and explain everything to us :-)

  • http://profile.typekey.com/robinhanson/ Robin Hanson

    there is no evidence that could ever serve to show that we are _not_ living in a simulation.

    James, I pointed to evidence that would make the theory substantially more likely, and any evidence that makes a theory more likely also makes it less likely that the theory is false. I think what you want is a possibility of definitive evidence that, if observed, would make it virtually certain that the theory was false. I agree that such evidence is not obviously forthcoming in this case. But I don’t see why we should insist on such a possibility.

  • http://www.stafforini.com Pablo Stafforini

    James, you also seem to endorse a claim that has remained unchallenged throughout this discussion, namely that there is no evidence that could either confirm or disconfirm that God exists. This is simply false. To take just one example, the so-called evidential (or inductive) problem of evil is an attempt by atheists to argue against the theistic hypothesis on the grounds that the extent, intensity and distribution of the observed moral and natural evils disconfirm the proposition that there is a being who shares the divine attributes of omniscience, omnipotence and perfect benevolence. Many non-believers, myself included, are not theists at least partly because they find this argument persuasive.

  • http://profile.typekey.com/sentience/ Eliezer Yudkowsky

    James, if there is any experience that could make you think you are more likely to be living in a simulation, then your failure to observe such experiences is evidence against being in a simulation. If anything could shift your belief in one direction, there must be something that can shift your belief in the opposite direction, by Conservation of Expected Evidence.

    This sort of thing is why I pointed out the difference between rationality and science.

  • RIchard Hollerith

    One purpose I have seen writers give for a simulation is to provide subjective experiences for the inhabitants of the simulation. I realize that this is a minority position, but I assign no intrinsic value to anyone’s subjective experience (whether or not you live in a simulation). It is only objective reality that has intrinsic value or utility, in my philosophy. (Of course, this position has ramifications for many important questions other than whether we are living in a simulation.) If in fact we are living in a simulation whose purpose is to provide us with subjective experiences, then we are living in a world with a foolish or unimportant purpose, so it does not much matter what we do or what we decide.

    The only important purpose for running a simulation I can think of is to model objective reality. So, I believe that we should choose to believe that if we are in a simulation, its purpose is to model reality accurately because if that is not its purpose, then the simulation and by extension ourselves, are unimportant.

    When confronted with two possibilities, under one of which we have an important role to play and under the other of which we do not, we should alway assume that the former possibility is the only possibility. Per analogy with Occam’s razor, I call that the relevancy razor.

    A simulation whose purpose is to model reality that is not ruined by the simulation’s inhabitants becoming aware they are in a simulation is an unusual simulation, and I would not be surprised if that fact can be used to extend the analysis significantly, but this comment is already too long.

  • http://raghavgupta.wordpress.com Fractal Folds

    I’ve gone through all the arguments in favor of this theory, all the comparisons with the great discoveries (gravity,relativity etc), but please, can someone mention at least one shred of manner in which this knowledge will benefit mankind? With every other such ground-breaking proposition, there is usually at least one major gain, some outcome of the new knowledge which physically changes our world forever. I would like someone to articulate what exactly is that proposed gain here?

  • Carl Shulman

    “If in fact we are living in a simulation whose purpose is to provide us with subjective experiences, then we are living in a world with a foolish or unimportant purpose, so it does not much matter what we do or what we decide.”
    Richard,

    Why is living in a world created with a foolish or unimportant purpose so lacking in meaning relative to a world with no purpose at all, e.g. a ‘basement-level’ universe not created by an intelligence? If a human being is conceived by parents using IVF with the purpose of enabling the child to live a happy life and have certain subjective experiences, does that make the child’s life pointless?

  • http://julesandjames.blogspot.com/ James Annan

    “I agree that such evidence is not obviously forthcoming”

    But do you agree that such evidence is obviously not forthcoming, and never could be? Maybe the possibility of falsification (indeed, inevitability of falsification at some level) is a slightly naive and idealistic viewpoint as to what science is all about, but there is, I think, a nugget of value in the principle.

    I’m not really interested (in this thread) in debating the value of non-black non-ravens or whatever in updating our Bayesian beliefs about whether “this sentence is false” is true or false. I’m a firm believer in Bayesian probability as a useful tool for many problems, but last I looked, it wasn’t the law.

    Does no-one have anything to say about my “creation argument”, and why Nick Bostrom’s proposal is any more interesting (other than that people here get very excited about technofuturism)?

  • http://profile.typekey.com/sentience/ Eliezer Yudkowsky

    James, the reply to your “creation argument” is that running simulations is a straight-line extrapolation from our modern models of physics; while becoming supernatural is an ill-formed concept which to the extent that it says anything at all, violates not only our modern laws of physics but also the monistic, mathematically simple character of physical law. (A simulation like ours could only run on a simulator capable of reliably processing mathematically simple systems, meaning it had enough stability to simulate a stable universe.)

  • Richard Hollerith

    Carl Shulman,

    Happiness has high instrumental value, so even a person who assigns no intrinsic value to happiness will tend to spend much time and effort on happiness — of himself and those around him. Let us take a little time to describe this instrumental value.

    Certain emotional states such as social confidence, happiness and love conduce to doing valuable and important work in the modern environment.

    Moreover, if a person fails regularly to have happy experiences, his ground state will tend over time toward sadness. Sadness and anxiety (and probably embarassment and shame) are no longer useful emotions in almost all situations. To avoid prolonged sadness and other unproductive emotional ground states, most people are well advised intentionally to seek out regular happy experiences.

    Although certain social situations (e.g., involving authority figures or groups with norms against “negativity”) can cause a person to feel happier than his physiology warrants, in general happiness closely tracks health and nothing makes it harder to feel happy than poor health. Health of course has very high instrumental value.

    To round out our description, we ought to note that happiness also tracks certain parameters of the interpersonal environment that resulted in high reproductive fitness in the environment of evolutionary adaptation: e.g., whether the people with whom you interact respect you and take your needs and wants into account. This is what psychologists mean by “social support system”.

    One of the most distressing aspects of the current situation is that most people seem perfectly happy with happiness as their ultimate good and see no need to make a sustained intellectual search for values that might trump happiness. John David Garcia (whose books are at see.org) focused on this problem.

  • Richard Hollerith

    Carl Shulman asks, “Why is living in a world created with a foolish or unimportant purpose so lacking in meaning relative to a world with no purpose at all, e.g. a ‘basement-level’ universe not created by an intelligence?”

    Since I reject happiness as an intrinsic value, the relevancy razor (described above) forces me to search for another intrinsic value. I believe that the search should be conducted using essentially the same methods we would use to refine our understanding of the physical environment (namely, physics and cosmology) and that it is possible for the search to end in success. That is, I do not hold to the is-ought distinction that most writers hold to, but rather believe in “objectively-valid morality” though I prefer to avoid the connotations of the word “morality” and to use instead “objectively-valid ultimate goal”. Although I am perfectly aware that people who believe they know the objectively-valid good have caused the slaughter of hundreds of millions of people, I see no choice but to continue the search just as I see no choice but for us to maintain and refine a model of our environment, that is, of what is. Moreover, I believe that the search is more than a matter of aesthetics, feeling, intuition or personal subjective opinion.

    In other words, I reject your suggestion that if ours world is not a simulation, it has no purpose.
    (I prefer the word “goal” to purpose. It has a sharper meaning.)

    John David Garcia wrote what I consider a good book-length description the objectively-valid good. I consider Eliezer’s Meaning of Life FAQ a good article-length description.

  • anonymous

    Now Scott Aaronson has written about this discussion over at Shtetl-Optimized:

    http://scottaaronson.com/blog/?p=265

  • http://www.hopeanon.typepad.com Hopefully Anonymous

    The idea that we’re in a simulated reality is a perhaps testable claim. As I’ve posted here, to go from “we may be in a simulated reality” to “maybe we should try to be as interesting as possible to our simulators” moves to wish fulfillment -one could as easily move in a direction such as “maybe we should try to be as inconspicuous as possible to our simulators” or “maybe we should keep doing exactly what we’re doing -it seems to be working”, etc. I usually find it annoying when a good idea is arbitrarily hitched to a not necessarily good idea. So, I think this topic should start with de-linking the hypothesis that we’re in a simulated reality from a particular behavior prescription that we “should try to be as interesting as possible”.

  • http://profile.typekey.com/nickbostrom/ Nick Bostrom
  • http://profile.typekey.com/sentience/ Eliezer Yudkowsky

    HA: I usually find it annoying when a good idea is arbitrarily hitched to a not necessarily good idea. So, I think this topic should start with de-linking the hypothesis that we’re in a simulated reality from a particular behavior prescription that we “should try to be as interesting as possible”.

    Seconded. It’s far more likely that our simulators are nothing like us, since they neither attempt to prevent Holocausts nor even attempt to facilitate them. The simulation, if it is one, seems extremely faithful to the notion of unified physical law; it is in an all probability an attempt to gain accurate knowledge about something that does run on unified physical law, or, perhaps, something that attaches intrinsic utility to patterns that are only “worthwhile” if they are accurate simulations of unified physical law.

  • http://www.hopeanon.typepad.com Hopefully Anonymous

    Eliezer, great comment IMO.
    I’m not against considering an anthropic principle behind our (perhaps simulated) reality, but I do think there’s probably evidence as strong that the simulation is created along the lines Eliezer proposes (about something related to “unified physical law”) and that we may be an incidental byproduct. This reality seems to be filled with stuff operating under unified physical law, not with humans. Of course that’s not proof positive of anything, but “unified physical law” does seem to be a more defining feature of our reality, rather than things that have relations to human aesthetics such as being “interesting” in a way we can relate to.

  • http://profile.typekey.com/nickbostrom/ Nick Bostrom

    Just for the record, Woit’s “critique” claimed that the FHI is funded by the Templeton Foundation. In fact, we have not received any funding from the Templeton Foundation. (But I’d be happy if we got some.)

    I agree with HA and Eliezer that we should avoid bundling these various claims about simulation. I would go further, and urge that we always remember the difference between the simulation argument and the simulation hypothesis. The simulation argument does not show, and was not intended to show, that we are living in a simulation. The simulation argument claims that a certain tripartite disjunction is true, and the simulation argument is one of the disjuncts. (It would be fair to say, however, that if the simulation argument is sound, then the simulation hypothesis is worth taking more seriously than we would otherwise have reason to do.)

  • http://profile.typekey.com/nickbostrom/ Nick Bostrom

    oops – the penultimate sentence should of course read: “… and the simulation *hypothesis* is one of its disjuncts”.

  • http://magicdragon.com Jonathan Vos Post

    Assuming that you’re the Robin Hanson who received his Ph.D in 1997 in social sciences from Caltech, then I apologize for the length of my prior comments.

    “Jonathan and nc, you could have made your basic point in a few paragraphs.”

    We were educated to cite our sources. Once that’s done, it’s easy to be terse, as the cited article can be read in detail by the curious.

    You and I have also both critiqued the Doomsday Argument (mine in a submitted article to Analog, which was not bought, but excerpts from which were scattered over several issues).

    I also cited you in a paper at last year’s NKS Wolfram Conference, on whether or not one could enumerate all possible economic systems. I think I quoted you (with attribution) on Economics of Abundance.

  • http://profile.typekey.com/robinhanson/ Robin Hanson

    Yes we should not assume we know too much about simulator motives (if any), but that doesn’t mean we have no idea whatsoever of them. A large fraction of simulations run now are for entertainment purposes, so that seems a plausible thing to assume about the future. The fact that our physics seems reliable is not much evidence about that – as creatures in an entertainment simulation should also think that.

  • http://profile.typekey.com/sentience/ Eliezer Yudkowsky

    A larger fraction of simulations run now are for entertainment purposes, so that seems a plausible thing to assume about the future.

    Then we should expect the world around us to appear optimized for entertainment to a far greater degree, no? Consider “learning” vs. “entertainment”. Only the former predicts that reality will be as faithful as possible to a low-level physics, because it is an attempt to learn about an outside phenomenon that runs on low-level physics.

    Or to put it more simply: This, our world, may or may not be a simulation by some ahumane entity to learn about the initial conditions of Singularities it could run into; the Holocaust is faithfully simulated because it is what would happen under unified physics. But it’s probably not there for entertainment, unless the entity is maximally entertained only by faithful simulations of unified physics.

  • http://profile.typekey.com/robinhanson/ Robin Hanson

    Eliezer, I agree that a simulation run for entertainment purposes would pay less attention to getting the physics right. But, it would pay sufficient attention to make sure that the people in that world believe that their physics is exactly right. So the fact that we believe that our physics is right is not evidence that we are not in an entertainment simulation.

  • http://geniusnz.blogspot.com GeniusNZ

    If bostrom’s paper is published in a top jornal and so widly discussed why is it that I was able to cripple it with a few seconds of analysis? Is philosophy in that poor a state?

    Bottom line is the formula as shown doesn’t logically follow from its own foundations on the indifference principle (or DD argument). Easy enough to demonstrate if anyone cares although its pretty straight forward.

  • http://profile.typekey.com/tim_tyler/ Tim Tyler

    Re: “This reality seems to be filled with stuff operating under unified physical law, not with humans.”

    You need a big universe just to get humans.

    Also, humans might be a step on the way to filling the universe with computronium – in which case all the inanimate matter observed might be seen as being the result of the simulation having only recently been booted up.

  • Collin

    The point is not whether the Simulation Argument “is” science, because that’ll never get an answer. The point is that readers do have a right to refuse to consider certain topics, or otherwise there would be no need for subject labels. And the Simulation Argument is definitely not what the label of science connotes in the science culture.

    As far as faith being useful to complement science, sometimes it is. Collapse Theory, for example, is a useful faith, because it makes some sense of quantum mechanics. I can’t see how the Simulation Argument can help anything except Matrix fan forums.