Uncommon Descent Serving The Intelligent Design Community

Following up Bostrom’s argument from simulation of universes . . .

Share
Facebook
Twitter
LinkedIn
Flipboard
Print
Email

That is, why inferring design on functionally specific, complex organisation and associated information, e.g.:

abu_6500c3magand equally:

cell_metabolism

. . . makes good sense.

Now, overnight, UD’s Newsdesk posted on a Space dot com article, Is Our Universe a Fake?

The article features “Philosopher Nick Bostrom, director of the Future of Humanity Institute at Oxford University.”

I think Bostrom’s argument raises a point worth pondering, one oddly parallel to the Boltzmann brain popping up by fluctuation from an underlying sea of quantum chaos argument, as he discusses “richly detailed software simulation[s] of people, including their historical predecessors, by a very technologically advanced civilization”:

>>Bostrom is not saying that humanity is living in such a simulation. Rather, his “Simulation Argument” seeks to show that one of three possible scenarios must be true (assuming there are other intelligent civilizations):

  1. All civilizations become extinct before becoming technologically mature;
  2. All technologically mature civilizations lose interest in creating simulations;
  3. Humanity is literally living in a computer simulation.

His point is that all cosmic civilizations either disappear (e.g., destroy themselves) before becoming technologically capable, or all decide not to generate whole-world simulations (e.g., decide such creations are not ethical, or get bored with them). The operative word is “all” — because if even one civilization anywhere in the cosmos could generate such simulations, then simulated worlds would multiply rapidly and almost certainly humanity would be in one.

As technology visionary Ray Kurzweil put it, “maybe our whole universe is a science experiment of some junior high school student in another universe.”>>

In short, if once the conditions are set up for a large distribution of possibilities to appear, you have a significant challenge to explain why you are not in the bulk of the possibilities in a dynamic-stochastic system.

Let me put up an outline, general model:

gen_sys_proc_modelSuch a system puts out an output across time that will vary based on mechanical and stochastic factors, exploring a space of possibilities. And in particular, any evolutionary materialist model of reality will be a grand dynamic-stochastic system, including a multiverse.

Now, too, as Wiki summarises, there is the Boltzmann Brain paradox:

>>A Boltzmann brain is a hypothesized self aware entity which arises due to random fluctuations out of a state of chaos. The idea is named for the physicist Ludwig Boltzmann (1844–1906), who advanced an idea that the Universe is observed to be in a highly improbable non-equilibrium state because only when such states randomly occur can brains exist to be aware of the Universe. The term for this idea was then coined in 2004 by Andreas Albrecht and Lorenzo Sorbo.[1]

The Boltzmann brains concept is often stated as a physical paradox. (It has also been called the “Boltzmann babies paradox”.[2]) The paradox states that if one considers the probability of our current situation as self-aware entities embedded in an organized environment, versus the probability of stand-alone self-aware entities existing in a featureless thermodynamic “soup”, then the latter should be vastly more probable than the former.>>

In short, systems with strong stochastic tendencies tend to have distributions in their outcomes, which are dominated by the generic and typically uninteresting bulk of a population. Indeed this is the root of statistical mechanics, the basis for a dynamical understanding of thermodynamics i/l/o the behaviour of large collections of small particles.

For instance, one of my favourites (explored in Mandl) is an idealised two-state element paramagnetic array, with atoms having N-pole up/down, a physical atomic scale close analogue of the classic array of coins exercise. We can start with 500 or 1,000 coins in a string, which will of course pursue a binomial distribution [3.27 * 10^150 or 1.07*10^301 possibilities respectively, utterly dominated by coins in near 50-50 outcomes, in no particular orderly or organised pattern], then look at an array where each atom of our 10^57 atom sol system has a tray of 500 coins flipped say every 10^-13 – 10^-15 s:

sol_coin_fliprThe outcome of such an exercise is highly predictably that no cases of FSCO/I (meaningful complex strings) will emerge, as the number of possible observed outcomes is so small relative to the set of possibilities that it rounds down to all but no search, as the graphic points out.

This is of course an illustration of the core argument to design as credible cause on observing FSCO/I, that once functionally specific complex organisation and associated information are present in a situation, it demands an observed to be adequate explanation that does not require us to believe in statistical needle- in- vast- haystack- search- challenge miracles:

islands_of_func_challAlso:

is_ o_func2_activ_info

The Captain Obvious fact of serious thinkers making similar needle in haystack arguments, should lead reasonable people to take pause before simply brushing aside the inference to design on FSCO/I. Including in the world of life and in the complex fine tuned physics of our cosmos that sets up a world in which C-chemistry, aqueous medium terrestrial planet life is feasible.

But we’re not finished yet.

What’s wrong with Bostrom’s argument, and wheere else does it point.

PPolish and Mapou raise a point or two:

>>1

  • Simulated Universes scream Intelligent Design. Heck, Simulated Universes prove Intelligent Design.

    I can see why some Scientists are leaning in this direction. Oops/Poof does not cut it any more. Unscientific, irrational, kind of dumb.

  • ppolish,

    It’s a way for them to admit intelligent design without seeming to do so (for fear of being crucified by their peers). Besides, those who allegedly designed, built and are running the simulation would be, for all intents and purposes, indistinguishable from the Gods.

    Edit: IOW, they’re running away from religion only to fall into it even deeper.>>

 

In short, a detailed simulation world will be a designed world.

Likewise High School student projects do not credibly run for 13.7 BY. Not even PhD’s, never mind Kurzweil’s remark.

So, what is wrong with the argument?

First, an implicit assumption.

It is assuming that unless races keep killing off themselves too soon, blind chance and mechanical necessity can give rise to life then advanced, civilised high tech life that builds computers capable of whole universe detailed simulations.

But ironically, the argument points to the likeliest, only observed cause of FSCO/I, design, and fails to address the significance of FSCO/I as a sign of design, starting with design of computers, e.g.:

mpu_modelWhere, cell based life forms show FSCO/I-rich digital information processors in action “everywhere,” e.g. the ribosome and protein synthesis:

Protein Synthesis (HT: Wiki Media)
Protein Synthesis (HT: Wiki Media)

So, real or simulation, we are credibly looking at design, and have no good empirical observational grounds to infer that FSCO/I is credibly caused by blind chance and mechanical necessity.

So, the set of alternative possible explanations has implicitly questionable candidates and implicitly locks out credible but ideologically unacceptable ones, i.e. intelligent design of life and of the cosmos. That is, just maybe the evidence is trying to tell us that if we have good reason to accept that we live in a real physical world as opposed to a “mere” speculation, then that puts intelligent design of life and cosmos at the table as of right not sufferance.

And, there is such reason.

Not only is it that the required simulation is vastly too fine grained and fast-moving to be credibly  centrally processed but the logic of complex processing would point to a vast network of coupled processors. Which is tantamount to saying we have been simulating on atoms etc. In short, it makes good sense to conclude that our processing elements are real world dynamic-stochastic entities: atoms, molecules etc in real space.

This is backed up by a principle that sets aside Plato’s Cave worlds and the like: any scheme that implies grand delusion of our senses and faculties of reasoning i/l/o experience of the world undermines its own credibility in an infinite regress of further what if delusions.

Reduction to absurdity in short.

So, we are back to ground zero, we have reason to see that we live in a real world in which cell based life is full of FSCO/I and the fine tuning of the cosmos also points strongly to FSCO/I.

Thence, to the empirically and logically best warranted explanation of FSCO/I.

Design.

Thank you Dr Bostrom for affirming the power of the needle in haystack challenge argument.

Where that argument leads, is to inferring design as best current and prospective causal explanation of FSCO/I, in life and in observed cosmos alike. END

Comments
EDTA, some of that is mathematics. KFkairosfocus
August 9, 2015
August
08
Aug
9
09
2015
01:02 PM
1
01
02
PM
PDT
Thank you for your reply, but I don't quite follow the logic all the way to the conclusion. It seems that to argue against the idea that we're running in a simulation, we have to invoke limitations on information processing (control theory, etc.) from _our_ universe. But how do we establish that computation/control in another realm is bound by the limitations of our universe? (Even Bostrum makes some assumptions like this without justifying them.) All I can see us concluding is that if we're simulated, then our universe is necessarily less fast/expansive/capable/continuous/whatever than the realm where the simulation lives. IOW, I agree, KF, that you've made your point _if_ we can assume that other realm(s) are in fact bound by limits on computation and physics that we have in our universe. But with precious little information regarding heaven (the only other realm I actually acknowledge), how can we make these assumptions? I'm not trying to side with those saying that there are multiple levels of simulation going on, or that we will someday be simulating universes here. I think we will hit limits on computation time/space long before being able to run usefully large simulations. I'm actually asking a much more modest question about what we can be certain of. Let me try to phrase it differently: The wild idea from the far side is that we might exist inside of a machine of some kind, and that our existence is contingent, but that we are designed. Our simulated universe thus does not exist independently of the simulator/designer. The designer can intervene at any time, observe anything it wants within, or just let things run their course, as desired. It could call off the simulation at any time. We are just 1's and 0's (or whatever the equivalent is in the higher realm—I don't even want to assume anything about the nature of information in a higher realm). On the other side are us theists, who say that a supreme Being designed the universe and us out of matter, but that our existence is contingent. God can intervene at any time, observe anything or everything He wants to, or just let things run their course. He can call things to a close with the sound of a trumpet at the time of His choosing and do whatever He pleases with us, including giving us new bodies of a different nature in His realm. So the crucial difference between those two scenarios is what? If you say, “Well, one is real, the other not,” then the question becomes, “How 'real' is a universe that might be very small in God's eyes, and that he can sweep away with just a word from His mouth?” And then we're right in the middle of thinking that we mean more as atoms, just because we can out-exist the 1's and 0's inside our computers. But that proves nothing. Scripture seems to indicate that we're as contingent and impermanent relative to God as our software is to us. I know the idea of a simulation universe is repulsive to us. But why is that so? Is it because it seems to make Him a liar, having supposedly set us in a universe that is quite old when it might not be? Only earthly simulations are limited to much less than 13 b.y., plus we can't assume that time exists outside our realm. Is it because we're making too much of analogies with our computers? Probably. Would it imply that He has deceived us in some other way, perhaps by letting us think we're more important and permanent than we really are? Whoa there! Hubris lies down that path.EDTA
August 9, 2015
August
08
Aug
9
09
2015
12:29 PM
12
12
29
PM
PDT
SA: general delusion trap, such self-undermines as it leads to breakdown of knowledge, rationality, ability to correct oneself etc. Self-refuting. KFkairosfocus
August 9, 2015
August
08
Aug
9
09
2015
05:40 AM
5
05
40
AM
PDT
ETDA
So I’m not taking the position that we are in a simulation, but rather saying that ruling it out looks tougher than you think. In particular, using features of our universe as the starting point doesn’t seem to work.
It's like Berkeley's argument that everything we experience is entirely subjective. We only know our own ideas of things and nothing really exists. It's impossible to refute that except that we generally don't live by it since we share a common experience of reality. It's also like Last Thursdayism.Silver Asiatic
August 9, 2015
August
08
Aug
9
09
2015
04:03 AM
4
04
03
AM
PDT
EDTA (& Mahuna): One may always pile speculation on speculation, creating a tower of seemingly persuasive fantasy. Indeed, that is how sci fi works. But that has little to do with reality. As a first problem, any world picture that implies general delusion (as opposed to particular, detectable errors) is self-undermining, leading in effect to the issue of an infinite regress of grand delusions. The classic example in many respects is a Plato's Cave world. But at least Plato had room for the prospective philosopher to detect the cheat, the apparatus of projecting the shadow-shows confused for reality. Video: https://www.youtube.com/watch?v=d2afuTvUzBQ&feature=related The computer simulation world is of course little more than yet another modern Plato's Cave, or a modern brain in a vat or Matrix dream pod world. Now, too, the matter is oddly connected to an updated form of von Mises' analysis of why Socialist centralised planning economics would fail. Information processing for a vast, distributed system. Now of course I appeal to physics, control theory and linked analysis of computation, some of it mathematical. If we take physics and linked analysis off the table in connexion with something that appeals to computation, we are building in a fundamental incoherence. A rule of thumb I long ago picked up for effective digital control (and, Mahuna an analogue system can be converted to an equivalent digital one) is that you want to sample 4 - 6 times in your fastest relevant rise time, setting sampling and processing requisites. In our world, the Planck time is of order 10^-44 s, and fast nuc/particle reactions may be down to 10^20 times slower. Fast chem rxns, ~ 10^30 times slower. To simulate such, you need to process smoothly enough to map a trajectory of interactions that run about that fast. With 4 - 6 samples per. Then, we have a cosmos of some 10^80 atoms, and likely a lot more than that photons in flight, etc. distributed across maybe 90 bn LY in the observable cosmos. And with a lot of simultaneous interactions at multiple scales up to galactic clusters and a connected cosmological expansion. Centralised processing to handle signals, algorithms and processing requirements to bring to one centre, process and redistribute, even if the scale of the sim is a lot smaller than the observed cosmos, would choke on the connectivity and processing required. To give an idea, two nodes have one link, three have three, four six, and so forth, the general mesh N = (1/2) * n*(n - 1). Connectivity is of order n^2. While a cosmos is not a full mesh all the way, that gives a fair idea of interactions to be handled. Take sol system ~ 10^57 atoms, with interactions that start with gravity, and implicate interconnexions lagged out to hours, but of order 10^114 just for gravity, and with a lot of physics and chemistry to be processed at rates ~ 19^15/s and 10^25/s. Choke. The far more effective topology to handle the distributed problem is a distributed network of loosely coupled, simpler processors that basically interact with near neighbours and allow global results to emerge from propagation. (This BTW is why economic planning by households and firms interacting through markets is more successful.) It soon emerges that the logical processor size is equivalent to atoms, molecules and particles interacting per what we would call physics and chemistry. And so, it makes sense to revert to, we live in a world based on such particles, laws and circumstances etc. An analogue world, i.e. a world of effective continuum. The grand simulation model has collapsed. But as a bonus, we see one reason to build a cosmos out from interacting particles, it allows nesting the design as you scale up to the cosmic scale and linked chunking and modularity. In effect instead of one grand global mesh, you have much smaller meshes that allow emergent meshes at higher levels. Laws and scales in a modular pattern with nesting. Linked, there is no empirically observed pattern of origin of FSCO/I by blind chance and necessity. Intelligent design is the observed adequate cause. So the assumption of emerging cvilisations with the level of simulation capability fails. Bostrom ends up only inadvertently underscoring the power of the needle in haystack challenge. And pointing to yet more compass needles pointing to design. We even see a good reason for the observed world nesting particles and atoms, molecules, bodies based on those and systems of interacting bodies up to cosmic scale. It is the logical way to conceive and design a relatively simple, robust world. KFkairosfocus
August 9, 2015
August
08
Aug
9
09
2015
12:11 AM
12
12
11
AM
PDT
I don't quite follow why the above rules out a simulation. It sounds as if we're saying that a (possibly immaterial) realm beyond ours cannot host a simulation of our universe (perhaps in computers of a nature we cannot conceive of), yet a Being can make a universe just as large and set it running (by means we do not understand). Arguments by reference to time, (e.g., by saying that the simulation has been running for 13 billion years), is meaningless, as there may not be such a thing as time in whatever realm the simulation runs in. And there weren't any intelligent observers here that far back either, to prove that it was really running back then. References to the number of atoms in our universe and why that many couldn't be simulated is ineffective, as it might not be necessary to simulate every atom at every instant. Nor do we know what another realm would be limited to re size. So I'm not taking the position that we are in a simulation, but rather saying that ruling it out looks tougher than you think. In particular, using features of our universe as the starting point doesn't seem to work.EDTA
August 8, 2015
August
08
Aug
8
08
2015
07:45 PM
7
07
45
PM
PDT
Nonsense. The people running the simulation know that THEY are not in the simulation because they're responsible for the care and feeding of the machines running the simulation. So, SOME of the beings in the universe that believe they are capable of free will MIGHT in fact be characters in a simulation run by a being that actually has free will. But the family and co-workers of the guys running the simulation can see the real physical universe. Unless the argument is really that it's not so much a simulation as a dream. And of course people have been suggesting that our lives are mere dreams of the Deity for millennia. Substituting "simulation" for "dream" is not a significant change. And the power of a being who can keep the same simulation (or dream) running for several billion years without interference from that being's family and co-workers qualifies the Simulator (or Dreamer) as what we commonly call "God". Also, modern nerds forget that before everything went digital, computers were analog. And so there is the very good chance that the universe we perceive is an analog simulation (at 1:1 scale) of some process we can't begin to imagine. (It helps if you've seen "The Hitch Hiker's Guide") But what is the difference between admitting that we're cogs in cosmic machine whose purpose we can never know and choosing to believe that that same cosmic machine is under "manual" control by the Deity? After all, heaven/paradise/the other world may also be part of the machine.mahuna
August 8, 2015
August
08
Aug
8
08
2015
02:56 PM
2
02
56
PM
PDT
SA, he is discussing an esoteric issue and posing it in terms of cosmology not directly disputing the vaunted mechanisms held to account for origin and diversification of biological life. Likely, he does not have "suspicious connexions." He is also based in the UK. And, there is the out that he is really addressing sustainability of civilisations, as in beyond a threshold they kill themselves off. That is close to the Drake equation and the issue of the great silence . . . we should hear somebody out there, unless, uh oh. KFkairosfocus
August 8, 2015
August
08
Aug
8
08
2015
11:23 AM
11
11
23
AM
PDT
IDist Nick Bostrom: “richly detailed software simulation[s] of people, including their historical predecessors, by a very technologically advanced civilization” He knows who the Designer is. I don't think he'll tell us "who designed the designer" but maybe he can find out. Bostrom can get away with this and not be considered pro-ID however, for some reason.Silver Asiatic
August 8, 2015
August
08
Aug
8
08
2015
10:22 AM
10
10
22
AM
PDT
REW
The thing is if Bostrom’s argument is compelling, than all of the evidence you think you have for ID and the existence of God can be subsumed within the alien simulation.
ID is falsifiable so another explanation could be more compelling. Additionally, Bostrom's argument is non-scientific. With that, there is an infinite number of non-scientific proposals that could counter ID, including a giant magic trick by some unknown alien created this universe.
All of these bits of evidence were just programmed into the simulation and probably have no bearing on reality.
Exactly. I could propose that the universe was created by a big pink unicorn. Then I could look at all the evidence in the universe that looks like a pink unicorn. Shapes, colors, movements ... I can also claim that "pink unicorns actually do various things". Thus, I'd have more evidence. If this was compelling to someone, then ID would not be successful for that person. But even easier, we don't need evidence. Someone can merely say that a simulation exists. That's it. No direct evidence for it is possible, and it is capable of any feature.
If his argument works we have zero evidence for God and 100% certainty that we’re in a simulation
The argument for God answers far more questions than the simulation does, and it has far more evidence in support of. The pink unicorn (multiverse/simulation/alien race) answers the existence of a universe. It does not answer the origin of the moral law, the origin of purpose in human life, the distinction between good and evil -- and a multitude of religion-specific evidences (miracles, prophecies, post-death events). But even still, someone can claim that the pink unicorn created humanity, gave it a purpose, gave it a moral law, provided a scale of value from evil to goodness to perfection and manifested miracles and prophecies. The pink unicorn is the first, non-contingent being and therefore possesses the fullness of being (which could not be added to or taken away from since there is nowhere for it to come from or go to) and is therefore all good and possessing the fulness of knowledge and power. That pink unicorn is starting to sound kind of familiar. :-)Silver Asiatic
August 8, 2015
August
08
Aug
8
08
2015
10:19 AM
10
10
19
AM
PDT
Hi REW, I am not endorsing Bostrom, in fact I go on to highlight key ways in which the simulation idea falls apart. The key point is that he acknowledges the relevance of seeing where the bulk of a statistical distribution lies and implications of a needle in haystack search. That is a point that has been dismissed, derided and hyperskeptically objected to for years. But here we have Bostrom and Boltzmann acknowledging it and using it as a matter of course. Let's clip:
Not only is it that the required simulation is vastly too fine grained and fast-moving to be credibly centrally processed but the logic of complex processing would point to a vast network of coupled processors. Which is tantamount to saying we have been simulating on atoms etc. In short, it makes good sense to conclude that our processing elements are real world dynamic-stochastic entities: atoms, molecules etc in real space. This is backed up by a principle that sets aside Plato’s Cave worlds and the like: any scheme that implies grand delusion of our senses and faculties of reasoning i/l/o experience of the world undermines its own credibility in an infinite regress of further what if delusions. Reduction to absurdity in short. So, we are back to ground zero, we have reason to see that we live in a real world in which cell based life is full of FSCO/I and the fine tuning of the cosmos also points strongly to FSCO/I. Thence, to the empirically and logically best warranted explanation of FSCO/I. Design. Thank you Dr Bostrom for affirming the power of the needle in haystack challenge argument.
So, dismissers, it is time to rethink. KFkairosfocus
August 8, 2015
August
08
Aug
8
08
2015
08:47 AM
8
08
47
AM
PDT
I'm not sure why an ID advocate would promote ideas such as this because I can't imagine that any IDers would actually believe it. Its almost as if you're presenting a choice to atheists: 'you can either believe what we believe about God creating the universe etc etc or you have to accept that we're just an alien simulation. The thing is if Bostrom's argument is compelling, than all of the evidence you think you have for ID and the existence of God can be subsumed within the alien simulation. All of these bits of evidence were just programmed into the simulation and probably have no bearing on reality. If his argument works we have zero evidence for God and 100% certainty that we're in a simulation ( I don't believe his argument works.....of course...)REW
August 8, 2015
August
08
Aug
8
08
2015
08:13 AM
8
08
13
AM
PDT
Bostrom's argument from simulation of universes has some surprising onward implications that support design theory.kairosfocus
August 8, 2015
August
08
Aug
8
08
2015
05:53 AM
5
05
53
AM
PDT

Leave a Reply