When Does the Universe Compute?

KentuckyFC writes "The idea that every physical event is a computation has spread like wildfire through science. That has triggered an unprecedented interest in unconventional computing such as quantum computing, DNA computing and even the ability of a single-celled organism, called slime mold, to solve mazes. However, that may need to change now that physicists have worked out a formal way of distinguishing between systems that compute and those that don't. One key is the ability to encode and decode information. 'Without the encode and decode steps, there is no computation; there is simply a physical system undergoing evolution,' they say. That means computers must be engineered systems based on well understood laws of physics that can be used to predict the outcome of an abstract evolution. So slime mold fails the test while most forms of quantum computation pass."

Comments

Joke of the Day

[Anonymous Coward]

"I'll get it," a wife said to her husband as the phone rang.
On the line a pervert, breathing heavily, said, "I bet you have a tight asshole with no hair."
"Yes," she responded. "He's sitting next to me watching TV."

Re:Joke of the Day

[Hognoxious]

More entertaining than the article, and contained more useful information.

Excellent blog, would buy again!

Re:Joke of the Day

[oldhack]

You know, this AC first-post joke is likely the best that can be salvaged from this bullshit "story".

Re:Joke of the Day

[Anonymous Coward]

I bet you waited all day for a chance to slip it in here.

Re:Joke of the Day

[Anonymous Coward]

....that's what she said.

Will it blend?

[marcello_dl]

Before we go into philosophy, I warn that being a computation is different from "being the result of a computation". The mold can be simulated.

Re:Will it blend?

[ackthpt]

Psst! It's an Analog Computer, they're way faster and more flexible than digital ones, but don't tell anyone or I'll be modded down into oblivion.

Re:Will it blend?

[TsuruchiBrian]

Analog computers are more flexible? Did you post this comment from an analog computer? Can I play candy crush on an analog computer?

How are you defining flexibility?

Re:Will it blend?

[Anonymous Coward]

Are you digital? (Aside from your fingers?)

Re:Will it blend?

[Anonymous Coward]

I think the OP's point is that you are an analog computer, and you can implement and play candy crush, which is clearly superior to being able to compute the game itself.

I don't think encoding/decoding are fundamental

[presidenteloco]

Why can we not think of the information as being embodied in "some aspect or other of" the matter and energy undergoing evolution. It is only some observer that needs to see the information as having been encoded or decoded.

Metrics of computations, or measurements of information flows, may be a productive way of describing (and predicting) complex physical evolutions, regardless of whether the physical system itself is identifiably encoding and decoding information explicitly. You just have to establish your own observer convention for how you think the information is represented in the matter and energy under discussion, or you can even just think about "the maximum amount of information" that could be contained in that matter/energy/spacetime region, and the maximum possible amount of information flow there.

Re:I don't think encoding/decoding are fundamental

[Xaedalus]

If I understand your argument, you're speculating that the universe is a pantheistic, evolving computation seeking entropy?

Re:I don't think encoding/decoding are fundamental

[Hatta]

We only don't see the encoding and decoding steps because we are inside the system that is doing the computation. If the universe were a simulation, those inside the simulation would see a ball trace a parabola with no encoding or decoding steps. Those who designed the simulation would be well aware of those steps.

Re:I don't think encoding/decoding are fundamental

[Codifex Maximus]

I used to adhere to the Theory of Entropy. Made sense at the time kinda like the flat world and Earth being the center of everything.

With a greater understanding of Physics and Cosmology, I've come to realize that the Earth is not flat nor the center of everything. With the realization that Energy is neither created nor destroyed only converted - the Theory of Entropy is disrupted in my mind and therefore proven false.

Now, discounting Entropy doesn't mean I deny that systems have a tendency to reach a Steady State where a perceived equilibrium has been established.

Re:I don't think encoding/decoding are fundamental

[camperdave]

Balls don't trace parabolas. They trace ellipses where one of the focal points is the center of mass of the Earth. In order to trace a parabola, the ball would have to be travelling at around 11.2 km/s

Re:I don't think encoding/decoding are fundamental

[Belial6]

I had this argument in another thread just yesterday. The laws of thermodynamics are obviously wrong. Wrong in the same way that Newtonian physics is wrong. Meaning that it is close enough for anything I will ever get my hands on, but that it clearly does not explain everything that is happening, and it is clearly violated at some point.

Re:I don't think encoding/decoding are fundamental

[Anonymous Coward]

Why can we not think of the information as being embodied in "some aspect or other of" the matter and energy undergoing evolution. It is only some observer that needs to see the information as having been encoded or decoded.

Sure, you can just sit on your ass forever while waiting for a system that happens to model the calculation you want and happens to be in the right state to match the initial conditions you wanted happens to fall into your lap. Have fun with that.

Re:I don't think encoding/decoding are fundamental

[Ultra64]

Balls don't trace parabolas.

Maybe not *your* balls.

Re:I don't think encoding/decoding are fundamental

[Anonymous Coward]

Those who designed the simulation

I know you meant this as an illustration, but I think it illuminates the flawed reasoning here. The ball's path is emergent from the rules of the simulation, there need not be any analytical solution or "designers". This is still computation by my definition.

Re:I don't think encoding/decoding are fundamental

[Artifakt]

When you try to define an observer in Physics, you run into a property of Quantum Mechanics. There, an observer is, for example, any outside particle that becomes involved in a state vector. When the state vector begins to impact what the particle is doing, that "collapses the vector", and the quantum state, which is until then is theoretically only a probablity, becomes an actual event, from the 'perspective' of that particle. You can also describe this process in terms of fields instead of particles, which still has much the same implications. Either way, you get an infinte regress, as it is possible to define a new state vector that includes that particle, and collapse that vector as it interacts with another particle, and so on.
          If you try to define an observer only as being capable of making some sort of interpretation, i.e., (from your own examples), by doing a prediction or applying a convention, you get an interesting problem. Schrodenger described this in the famous Cat Paradox, but let's scale it up a bit. If an observer has to be able to think, even very simplistically, then we can't logically ask what Jupiter is. I can, for example, say that, about an hour ago, Jupiter was a planet. because I looked up and saw it, and was capable of thinking about the difference between a planet and other things that look like one. But Jupiter, right now, isn't any actual thing, it's an uncollapsed state vector involving umpteen bajillion fundamental particles and a relativistic light cone about an hour long, and all of that becomes just a probability and not an actuality. If there's limitations like you have suggested, then any time it's cloudy everywhere a moderately educated person can see Jupiter and identify it, the state vector lengthens until the rain goes away, potentially becoming many hours long. That's one damned big 'cat', in an even bigger 'box'. Schrodinger figured a macroscopic object like a cat was big enough to show how some arguments were absurd.
            By this interpretation, every tree that falls in a forest with no observer around not only doesn't make a sound, it's an uncollapsed state vector until something which knows the difference between a tree and a shrub comes along. Does an insect tasting the fallen tree need to know that its a dead tree and not some other cellulose object before the state vector collapses? If we are the only astronomical civilization in the universe, did all those objects in the Hubble deep field photo just become objects when Hubble took the shot, or even when some human first looked at the picture? All the universe that lies outside 'the' observable radius could be one gigantic uncollapsed quantum probability function, not yet a thing, and since it is moveing away to fast for light to ever reach us, never to be an actual thing.
          All that's implied if encoding and decoding were fundamental, which Is probably why you don't like that interpretation much. If encoding and decoding are fundamental, that pretty much implies a "Big G" god to take the whole universe from probabilistic function to actuality, by observing from every point, or else a vast epistemological limbo surrounds our tiny bubble of observations. We can justify claiming that mathematical calculation based models describe a lot of events in a way that makes them easier to address, and such, as you've suggested, but if we claim to know for a fact that these reflect a fundamental truth, we'e claiming to know that something omnescient exists, and that's probably not a place a lot of the 'universe as computation' theorists want to go.

Re:I don't think encoding/decoding are fundamental

[HybridST]

When thermodynamics is wrong, you've missed something.

Re:I don't think encoding/decoding are fundamental

[Hatta]

Right, the "simulation" could be any algorithm, there needent be designers at all. We could all be living in the 10^10^10^10th iteration of Wolfram's Rule 30.

Re:I don't think encoding/decoding are fundamental

[Anonymous Coward]

aren't all the problems due to trying to apply micro concepts to the macro world? or the other way 'round?

Re:I don't think encoding/decoding are fundamental

[Anonymous Coward]

Well, I advanced through all of those stages of thought and am currently three steps ahead, so you should just abandon your current theories that improve on the backward theories pervasive among scientists who claim to study this stuff.

Re:I don't think encoding/decoding are fundamental

[K. S. Kyosuke]

The relative error of a common ball's trajectory, thrown at some random angle at speed common for a human arm, barring the aerodynamic drag, with regards to its actual trajectory, is roughly comparable to the relative error of your "11.2 km/s" with regards to actual local parabolic speed near Earth's surface.

Re:I don't think encoding/decoding are fundamental

[Anonymous Coward]

Since the designer of Wolfram's Rule 30 is Stephen Wolfram, I'm very confused by what you're trying to say.

Re:I don't think encoding/decoding are fundamental

[Anonymous Coward]

dude, throw a ball in the air it traces (pretty close to) a parabola.
y=yo + Vot - 1/2 g T ^2
i think you are putting a ball in orbit.

Re:I don't think encoding/decoding are fundamental

[Hatta]

One could easily argue that those rules were discovered, not designed. All Wolfram really did was design a numbering system.

Re:I don't think encoding/decoding are fundamental

[Anonymous Coward]

Thermodynamics is wrong? Have you found systems A, B, and C where A and B have the same temperature, B and C have the same temperature, but A and C don't? (That would mean the zeroth law is wrong.) Have you found a violation of conservation of energy? (That's the first law.) Has someone built an engine that can extract work from a single thermal reservoir? (That's the second law.) Do you know of a crystal at absolute zero with nonzero entropy? (That's the third law.)

I understand that scientific theories are usually approximations, but you're saying that thermodynamics is obviously wrong, and as a student of thermodynamics I would be very interested to hear your obvious argument.

Re:I don't think encoding/decoding are fundamental

[Anonymous Coward]

It's only statistical. It doesn't capture what's really going on. See the Fluctuation theorem.Entropy is not bad, we zip files creating more entropy to send them faster. Disorder is not bad, a lot of humor relies on "you didn't see that coming".

Re:I don't think encoding/decoding are fundamental

[Anonymous Coward]

Any object that is falling at a speed less than escape velocity is in an "orbit." It's just that when a person throws a ball, usually that orbit intersects the surface of the Earth.

Re:I don't think encoding/decoding are fundamental

[ultranova]

The laws of thermodynamics are obviously wrong. Wrong in the same way that Newtonian physics is wrong. Meaning that it is close enough for anything I will ever get my hands on, but that it clearly does not explain everything that is happening, and it is clearly violated at some point.

Please give an example of such violation? Because I'm afraid I can't see this obvious flaw you posit.

Re:I don't think encoding/decoding are fundamental

[Dcnjoe60]

When thermodynamics is wrong, you've missed something.

If thermodynamics is correct, then a computational universe is wrong.

Re:I don't think encoding/decoding are fundamental

[Anonymous Coward]

half-ellipses are pretty close to a parabola

Re:I don't think encoding/decoding are fundamental

[Belial6]

Our existence is an example of such violation.

Re:I don't think encoding/decoding are fundamental

[Belial6]

Because Magic!

To have a lower entropy state, you must have gotten there somehow. Humanity doesn't know how that happened, so scientists do what everyone else does when they are not even close to understanding something. They declare Magic!

Re:I don't think encoding/decoding are fundamental

[retchdog]

this is starting to sound interesting. could you elucidate please?

Re:I don't think encoding/decoding are fundamental

[jouassou]

Actually, he's right, and the analogy is quite good too. Newtonian physics is "wrong" in the sense that it doesn't hold for very massive, very fast or very small objects. However, for medium-sized objects moving at medium speeds, it holds very well.

Similarly, the second law of thermodynamics, that entropy always increases, can be derived in statistical mechanics by assuming that there are an infinite number of particles in your system. Thus, it holds for the entire universe, and it holds extremely well for any macroscopic system that I know of. However, for microscopic systems, it becomes quite probable that entropy decreases in small periods of time (the fluctuation theorem tells you the probability for this to happen.)

If you're interested in how this "makes sense": in statistical mechanics, it is shown that entropy is actually just a measure of microscopic disorder. There usually exists a lot more of possible disorderly states than orderly states for a system, so if no particular microstate is preferred (the probability of entering any microstate is equally probable), it's simply more probable that you will observe a transition from an ordered state to a disordered one, not the other way around. For a small system, the discrepancy is small, so you see transitions in both directions on small enough timescales. But as the number of particles in the system grows, the number of disordered states of the total system will grow far faster than the number of ordered states (the discrepancy is O(n!) for n particles in the system), so transitions from disordered to ordered states become extremely unlikely.

Re:I don't think encoding/decoding are fundamental

[Anonymous Coward]

Balls don't trace parabolas.

Maybe not *your* balls.

You better hope the rest of your equipment is going as fast as your balls, then. Otherwise... rip-ouch!

Re:I don't think encoding/decoding are fundamental

[TsuruchiBrian]

If you think the whole universe is a computer, then the phrase "I'm going to buy a new computer" is meaningless. There needs to be a differentiation between things that can be used for general computation, and other things (e.g. like the kind of computer that can only be used to compute how a particular slime mold will go through a particular maze at a particular time.

If you don't need the information to be encoded/decoded because you can just "see it" in there like the matrix, that information still represents something, and you are doing the encoding/decoding in your head. If a slime mold is actually doing a linked list insertion operation, you need to be able to say something like "See this little green blob over hear... The number of black spots on it represents the pointer to the next node". Even if you yourself do not require the information to be decoded, it needs to be able to be decoded for everyone else to be useful.

Re:I don't think encoding/decoding are fundamental

[Anonymous Coward]

In what way?

Re:I don't think encoding/decoding are fundamental

[TsuruchiBrian]

Nothing explains everything. Nothing *could* explain everything. Maybe our current theories of electromagnetism explain how systems behave by introducing concepts of protons and electrons, but they don't explain how protons and electrons have electrical charge to begin with. Even if we discover that protons have charge because of they have the charge of 2 up quarks and a down quark, it doesn't explain why the quarks have electrical charge.

This logically must lead either to an infinite regression or a dead end.

Re:I don't think encoding/decoding are fundamental

[crioca]

this is starting to sound interesting. could you elucidate please?

There's a number of things he could mean; but my guess is he's thinking of the digital model of a computational universe, which clashes with thermodynamics in a number of ways as opposed to the quantum model, which doesn't. (provided we figure out the whole quantum gravity debacle)

If you're interested I recommend you pick up Seth Lloyd's Programming the Universe. It's a great introduction to QM and quantum computing and is totally accessible for anyone with a high school education.

Re:I don't think encoding/decoding are fundamental

[Anonymous Coward]

I understand what you're saying and I've made this same argument before.

You could take any system. Take every state it was in from one point in time or space to any other point in time or space. And arbitrarily map each state to some stage of a computation.

Then the very nature of the system was that it was doing this computation.

In this way you could say that any fleck of space dust, in any moment or eon of time simulated our exact universe or any other.

Re:I don't think encoding/decoding are fundamental

[crioca]

In the first paragraph he's saying that it's unnecessary to think of information as encoded as matter and energy, because from the universe's perspective energy and matter is information and it's all being processed at the bit level. (although it's actually the information level, because it's not binary, so no bits involved)

In the second paragraph he's saying that measuring this information as it's processed could be useful as it would allow us to model physical interactions with near perfect accuracy, but we'd have to come up with our own way of encoding this information into format we can manipulate.

If I understand your argument, you're speculating that the universe is a pantheistic, evolving computation seeking entropy?

It's a mistake to think of the computational universe model as having agency. It's just the idea that since all activity in the universe can be abstracted as information (this is the foundation of quantum mechanics), then every interaction is a computation.

tl;dr: The universe laughs at your tiny minds and their ideas of magical superhumans and information that cannot be known. Ho ho ho.

Re:I don't think encoding/decoding are fundamental

[crioca]

I used to adhere to the Theory of Entropy. Made sense at the time kinda like the flat world and Earth being the center of everything.

With a greater understanding of Physics and Cosmology, I've come to realize that the Earth is not flat nor the center of everything. With the realization that Energy is neither created nor destroyed only converted - the Theory of Entropy is disrupted in my mind and therefore proven false.

Uh, there is no "Theory of Entropy", entropy is a phenomena described in Information Theory. Maybe next time you set out to disprove a theory, make sure it's a real theory first.

Re:I don't think encoding/decoding are fundamental

[crioca]

I had this argument in another thread just yesterday. The laws of thermodynamics are obviously wrong. Wrong in the same way that Newtonian physics is wrong. Meaning that it is close enough for anything I will ever get my hands on, but that it clearly does not explain everything that is happening, and it is clearly violated at some point.

The day the 2nd law of thermodynamics is proved wrong is the day one of Stalin's descendants goes back in time, kills Stalin, then returns to his own time and spends the rest of the day playing catch with his son in a world where Stalin never came to power.

Re:I don't think encoding/decoding are fundamental

[Dcnjoe60]

this is starting to sound interesting. could you elucidate please?

There's a number of things he could mean; but my guess is he's thinking of the digital model of a computational universe, which clashes with thermodynamics in a number of ways as opposed to the quantum model, which doesn't. (provided we figure out the whole quantum gravity debacle)

If you're interested I recommend you pick up Seth Lloyd's Programming the Universe. It's a great introduction to QM and quantum computing and is totally accessible for anyone with a high school education.

You would be correct. However, until somebody can reconcile quantum gravity, the quantum computing model of a computational universe is unworkable. As such, the only viable alternative would be the digital model, which conflicts with thermodynamics. BTW, the current quantum models can be made to work mathematically, but to do so requires accepting premises that would invalidate other accepted theories of QM. As a professor recently put it, it would require the cat to be both dead and alive while you are actually observing it. Since that can't be, unless somebody comes up with a new model, all that is left is the digital model.

Re:I don't think encoding/decoding are fundamental

[mcswell]

If you think you yourself compute, then I guess you must be paranoid--otherwise there'd be no observer decoding what you're doing. Right?

Re:I don't think encoding/decoding are fundamental

[Belial6]

Because one way or the other... "MAGIC!"

Re:I don't think encoding/decoding are fundamental

[Anonymous Coward]

This, from the same retards who think that graphing a pair singularities allows you to ignore the fundamental problem that you just DIVIDED BY ZERO. No thank you. I'll stick with good old Newtonian physics until you idiots come up with something that accurately explains the laws of motion AND magnetism.

Re:I don't think encoding/decoding are fundamental

[jouassou]

DIVIDED BY ZERO. No thank you. I'll stick with good old Newtonian physics until you idiots come up with something that accurately explains the laws of motion AND magnetism.

Sticking to Newtonian mechanics because the math looks prettier doesn't sound very scientific. Isn't the point of science to explain and predict the largest number of phenomena, with the smallest number of independent assertions?

Special relativity predicts and explains many phenomena that Newtonian physics doesn't, and even more so for general relativity. The same goes for quantum mechanics and quantum field theory, which can be used to derive all of chemistry and electromagnetism, and is the only theory so far that can predict what happens in particle accelerators. I'm not saying that any of these theories will never be superseded, but so far they explain a lot more than the theories we had in the 1800s.

Re:I don't think encoding/decoding are fundamental

[Your.Master]

To have a lower entropy state, you must have gotten there somehow.

Unproven assertion. The underlying assumptions are twofold:

1. That there is a point of minimum entropy, or a finite set of such points which all have an equally minimal entropy.
-- This has reasonably good evidence, but is hard to prove.
2. History must extend beyond the point of minimum entropy. By history, we mean that a "time" must exist "prior" to this state, where the Universe was in a different state, and it must not simply be looping through equivalent minimum entropy states.
-- The only support I see for this is argument from incredulity

Saying that people are declaring "magic" is just you being a dick. You are claiming some knowledge of perpetual motion that you cannot justify, you just pulling saying "magic" over and over in these threads. And you're claiming it's obvious, when it is no such thing.

I'll try to make a valid dichotomy (or trichotomy or n-chotomy, whatever it takes). One of the following must be true:

1. There was a "beginning" of some sorts, meaning a point of absolute minimum entropy for all time before which there was no previous state of higher entropy
2. There was no beginning, and the Universe "recycles" entropy in some sense (equivalently, Thermodynamics' Second Law is not absolute).
3. There was no beginning and no recycling, but the state of maximum entropy is unreachable -- it's either not finite, or in our asymptotic approach to the finite limit it's infinitely divisible. Eg. (as a fanciful example, not trying to be realistic) every black hole creates another Universe that may have lesser beginning entropy but also subdivides that entropy more finely so it contains a Universe of essentially the same complexity as the "parent" Universe for all intents and purposes.

For what it's worth, I actually am leaning toward the same conclusion you came to (option 2), but there's no way to justify it in the hardcore way you're justifying it, even as you wax poetic about how heat death is too strong a statement. I also suspect that #2 comes from random chance and is not harnessable. These are not things that I can support strongly -- what I can support strongly is that we don't have any good observational evidence that your suspicion is correct and the 2nd law is incorrect.

Re:I don't think encoding/decoding are fundamental

[Your.Master]

Where he's wrong in that post is primarily in the term "obviously", and secondarily in the justification for it that he thinks is so obvious (if you don't want to backtrace his posts, it's basically a variant of the argument from first cause, where in his case the "first cause" is some macroscopic entropy-reducing phenomenon being applied to a previous iteration of the Universe which had substantially higher entropy, and this process presumably repeats ad infinitum throughout an infinite history of the Universe).

Re:I don't think encoding/decoding are fundamental

[marcello_dl]

> there needent be designers at all

It's worse than that, there is no need for the concept of "there" "need" designer". That cuts both way because it invalidates most atheism reasoning as well. In fact religions "a god told me that..." are more logically acceptable than "if there were a god, then..."

> We could all be living in the 10^10^10^10th iteration of Wolfram's Rule 30
Yes... well... provided that the concept of "number" has any comparable meaning in the remaining 10^10^10^10-1 iterations. Once you get out this universe, you get out of the logic that "rules" it (because it doesn't rule anything, it merely models), unless you want to make a logic system a god ruling over all iterations, which is not even religion, it is idolatry.

Re:I don't think encoding/decoding are fundamental

[Anonymous Coward]

The difference between the parabola and the ellipse is orders of magnitude smaller than the difference of both to the actual trajectory of that ball, which is influenced by inhomogeneities in the earth's field, by the gravitation of sun and moon, by the Coriolis force due to the rotation of earth, by air friction, air drag, the Magnus effect, ...

Re:I don't think encoding/decoding are fundamental

[Anonymous Coward]

You didn't divide by zero, you simply reached the end of where you could speculate about. Okay, so you and I and the rest of the scientific community have this idea that you can't fit more than a certain amount of matter or energy into one spot -- at some point, gravity must fail to overcome some fundamental barrier. So to test this idea, we have to smash atoms together with really high energies to understand what happens to matter at that point.

The holes in Newtonian physics were wide enough to have a whole Theory of Relativity fit in them, and when we gather enough data about matter to discover (e.g.) why infinitely 'dense' points don't exist, the resultant understanding will likely be of equal impact to our understanding of the universe. Until then the physical theories based on Relativity will continue to be the best description of our observations of the universe.

Magnetism, really? What kind of a kook are you?

Re:I don't think encoding/decoding are fundamental

[presidenteloco]

There is a distinction between the quest to define what makes a useful computer, and the murky info-physics question(s) about whether fundamental aspects of physical evolution in the universe may be well-modelled in terms of information theory, chaitin-kolmogorov information theory, and computational theory.

Both are interesting questions. The latter info-physics speculations would probably require generalization of what can be considered to be "holding or embodying information" and generalization of what computing (information transformation) is considered to be. They would not benefit from restriction of what is considered to be computing or a computer.

At the edge of speculative and as yet unfounded info-physics pondering, is it just me, or is (quantity,flow of, localization/dispersion of) energy looking suspiciously like it should be modelled in terms of quantity, flow, localization/dispersion of information? Especially when it comes to understanding what "classical" physics is about, and what the transformation from quantum states to classical states may be about? I'll stop there. I know nothing and I know it well.

Re:I don't think encoding/decoding are fundamental

[TsuruchiBrian]

There is definitely information contained in every particle in the universe and it transforms in ways which are governed by physical laws.

There are some deterministic processes that can be used to computer lots of things (e.g. Turing machines). There are some deterministic systems that really can;t computer anything useful. Imagine a bit that flips between 0 and 1. You can change the input and get a different output, but the power of this "computer" is limited to computing the obvious.

Nature it seems is full of simple systems that are not useful, and complicated systems that are so complicated that we are not yet able to understand exactly how they work (e.g. human brains, or even ant brains).

I would categorize the whole universe as definitely too complicated to understand, if for no other reason than the number of particles that comprise the system. I would say it can be considered a computer, in that it contains everything including things known to be computers in it (e.g. a computer with stuff added to it is still a computer), but this is probably not how anyone else would look at it.

I think there is a distinction to be made between systems that we know are computers, and other systems which are definably too simple to be effective computers, and systems that are so complex that we don't know if they contain or are computers, and other systems which are known to be computers but we don't yet know how they work. I think this distinction should be made if for no other reason that practicality. I want to know which systems are known to be able to simulate other systems in a general way, and which we can;t or do not know how to use for that purpose.

Re:I don't think encoding/decoding are fundamental

[Anonymous Coward]

DIVIDED BY ZERO. No thank you. I'll stick with good old Newtonian physics until you idiots come up with something that accurately explains the laws of motion AND magnetism.

Sticking to Newtonian mechanics because the math looks prettier doesn't sound very scientific. Isn't the point of science to explain and predict the largest number of phenomena, with the smallest number of independent assertions?

Special relativity predicts and explains many phenomena that Newtonian physics doesn't, and even more so for general relativity. The same goes for quantum mechanics and quantum field theory, which can be used to derive all of chemistry and electromagnetism, and is the only theory so far that can predict what happens in particle accelerators. I'm not saying that any of these theories will never be superseded, but so far they explain a lot more than the theories we had in the 1800s.

Um, while I personally think it's a bad idea to ignore well-tested and effective theories that include unwanted features, it's hardly unscientific. New theories using different first principles is what drives science relentlessly forward. What is missing from the conversation, however, is a description of a working framework that has the "prettier math". This may be what the OP is developing, or suspects exists. But lack of a complete theory is not adequate justification for dismissal. Science has never produced complete theories.

Nonsensical sentence

[Anonymous Coward]

>The idea that every physical event is a computation has spread like wildfire through science. That has triggered an unprecedented interest in unconventional computing such as quantum computing, DNA computing and even the ability of a single-celled organism, called slime mold, to solve mazes

Quantum computing is no less physical than classical computing. DNA computing is entirely different than quantum computing. They have almost nothing in common. Is OP trying to say that physical models are used to perform computing? That is always the case. OP's hypothesis makes no sense.

Re:Nonsensical sentence

[Suiggy]

OP is referring to the paradigm shift of digital physics/pancomputationalism/holography where matter and energy are no longer fundamental, but rather emergent from information. However, the author's of the paper don't know what they're talking about, they're playing with semantics and dictionary definitions, trying to claim that things we don't understand at a macroscopic scale due to complexity aren't computation.

Well, I have a theory

[Anonymous Coward]

Point particles come from the interaction of various waves which carry force. Points don't even take up space. The Universe is one giant 2D wave, and all 3D space is holographic illusion.

Re:Well, I have a theory

[Anonymous Coward]

Point particles come from the interaction of various waves which carry force. Points don't even take up space. The Universe is one giant 2D wave, and all 3D space is holographic illusion.

Look, seriously, Doom was DECADES ago at this point. The rest of the world's moved on, you really need to stop living in the past and base your universal theories on something OTHER than an engine Carmack made back in the DOS era.

Re:Well, I have a theory

[Codifex Maximus]

To me, there are only four primary dimensions.

Length, Width, Height and Time.

The L*W*H dimensions ARE space. No need for a space dimension.

Time is a measurement used to delineate Frames.

I would say, do not limit yourself to only thinking in 2 dimensions as you cannot live in them. It takes 3. I will say that it also takes a bit more thought to see the world or understand it in 3 dimensions. One should try.

Re:Well, I have a theory

[Anonymous Coward]

http://gizmodo.com/our-universe-might-just-be-fourth-dimensional-black-hol-1410271260

Similiar idea but as a 3D wavefront in 4D space.

Re:Well, I have a theory

[K. S. Kyosuke]

The Universe is one giant 2D wave, and all 3D space is holographic illusion.

...and so are D cups? I feel cheated!

Re:Well, I have a theory

[hazah]

The holographic principle simply states that information of 3d space can be encoded onto a 2d surface. Likewise for 4d spacetime & 3d space. So any number of dimentions can, in fact, be represented within the confines of that number -1 dimentions. At least, according to the theory.

Probably meant "useful computation"

[Anonymous Coward]

The article claims that computation requires the theory behind it to be well-understood.
However, information processing is computation, whether you understand it or not.

That would be like saying that the brain did not compute because our theories behind how a brain works are insufficient to fully understand it in its entirety.

Definitions

[doconnor]

Sounds like they are arbitrarily defining computing as a simulation of the universe, therefore the actual universe cannot compute. I think this unnecessarily limiting people's imagination.

Besides, from the inside of a simulation its all real to you.

Re:Definitions

[gstoddart]

Sounds like they are arbitrarily defining computing as a simulation of the universe, therefore the actual universe cannot compute.

Or that they're saying the universe doesn't need to do calculations to determine where a falling object is going -- it just falls according to the laws of physics and doesn't need to be calculated.

I think this unnecessarily limiting people's imagination.

Does 'imagination' in this context actually tell us anything? We know that we need to do calculations for this stuff, but how does the assertion that the universe isn't doing the calculations limit our imagination? Stuff happens according to physical laws, the behavior is inherent to reality. Nobody has to do the math, it just happens.

Besides, from the inside of a simulation its all real to you.

Very meta, and equally meaningless. Yes, if we were in a simulation, we'd likely never know.

But given that we have no evidence to suggest we are, any assumptions around the notion that we are (or may be) are pretty much useless to us unless we can figure out the gaps in the simulation.

To me the suggestion we're living in a simulation serves no other purpose that throwing out something wacky to stump people at parties, but otherwise doesn't seem to have any application to understanding our universe.

Re:Definitions

[Obfuscant]

Or that they're saying the universe doesn't need to do calculations to determine where a falling object is going -- it just falls according to the laws of physics and doesn't need to be calculated.

Sounds like a Schroedinger's Cat of physical computation.

Hold a bowling ball over a cat. Drop the bowling ball. The cat is either a live or dead, the actual state cannot be determined until something decodes the computation done by the falling (or not falling) bowling ball.

"Damn, my printer jammed, we're overrun by half-dead cats." "We're half-overrun by dead cats?"

Re:Definitions

[Sarten-X]

doesn't seem to have any application to understanding our universe.

If we are a simulation, we may be able to discern exactly what we're simulating, and why. Theology aside, even discovering that our universe functions like a simulation may allow us to seek out and utilize aspects of the simulation that are useful to us.

Consider, for example, if we could simply access an information store outside our universe from anywhere within it. Even a single bit being accessible would offer the ability to have near-instant communications with other planets, or perhaps even other stars. If we could push matter out of or into the universe, we'd have an effective teleportation mechanism.

Science is all about figuring out the rules of our universe. Being inside a simulation means there are other, possibly different, rules outside, so breaking out means we have new capabilities that are impossible within our universe.

Re:Definitions

[marcello_dl]

> If we are a simulation, we may be able to discern exactly what we're simulating, and why.

No wai and I can prove it.

You have an mp3 player with two songs in it. The random playing algorithm makes it play the first song, the second song, the first again the second again and so on, because when it has to choose the next song there is only the other one available.

The normal playing algorithm plays the first song then the second song then the first song and so on EXACTLY LIKE the random playing algorithm.

Without looking at the mode, relying on your ears only, you would likely theorize that it's playing sequential, and some nerd would come up with the theory that it's playing random, some other nerd would argue about occam's razor and bullshit like that. In truth there is no way to discern the truth of the programming until you get to see the mode.

Think about it as plato's cave revisited.

Re:Definitions

[camperdave]

"Back in Nebraska, our cat got stuck in my brother's camp trunk, and we did not need to open it to know there was all kinds of dead cat in there." - Penny.

Re:Definitions

[gstoddart]

If we are a simulation, we may

I see ifs, maybes, what ifs, a few 'could bes', and wouldn't it be awesome if ... and nothing at all to suggest any of it is real or relates to our universe as we see and experience it.

Science is all about figuring out the rules of our universe.

Yes, yes it is.

It isn't about saying "gee, if our universe wasn't really a universe, then all of these things could be true and I could have a pony".

What you're describing is at best speculative fiction, and at worst just making shit up. Anything but science.

I'm not aware there's any credible evidence to suggest we are in a simulation which means everything after your first 'if' is pretty meaningless.

Science only deals in what we can see and observe, and that's not what you're describing. In fact, I'd say it's pretty much the opposite.

Re:Definitions

[Anonymous Coward]

It has as its application maintaining for consideration the notion of the universe as Designed, as the precise important reason, important in your very own mind, that you go to such strenuous argument to state that it isn't important.

The core Interpretations of QM (e.g. Copenhagen, Everett) do not have differentiating evidence for them either. Nor can we currently ascertain knowledge of what exists "beyond" the framework of possible existences they suggest. That doesn't mean we won't.

Re:Definitions

[UnknownSoldier]

> If we are a simulation, we may be able to discern exactly what we're simulating, and why.

As a mystic the reason is quite simple: The growth and evolution of consciousness.

There is one primary reason the [physical] universe exists: Relationships.

What, you thought human consciousness was the only kind !? There is NO-THING BUT consciousness.

See Peter Russel's excellent talk:

The Primacy of Consciousness - Peter Russell
http://www.youtube.com/watch?v=-d4ugppcRUE

Re:Definitions

[Artifakt]

I have absolutely no interest in learning if I exist inside a simulation or not, unless it's programed with accessable cheat codes.

Re:Definitions

[NoImNotNineVolt]

To me the suggestion we're living in a simulation serves no other purpose that throwing out something wacky to stump people at parties

... whoa... </neo>

Re:Definitions

[Sarten-X]

Your "proof" only illustrates one particular case, which does not necessarily apply to the question at hand.

It is indeed like Plato's cave, and the same lesson applies: Occam's razor is not always correct. Sometimes there are really weird truths, and with enough experiments, they can be discerned. Perhaps that MP3 player takes a little longer between songs while it tries to randomize the list. With a long enough recording of the playback at a high enough precision, the slight delay can be noticed, and that analyzing nerd can announce the mode with certainty.

The fundamental basis of all science is that we are wrong. Aristotle was wrong, Newton was wrong, Einstein was wrong, Hawking was wrong, and every scientist today is probably wrong about something. The scientific method allows us to become right. Each successive scientist makes more and better observations than his predecessors, notices what observations don't match the old law's predictions, theorizes new hypotheses to fit existing data, then experiments to test the hypothesis. Eventually, a new theory replaces the old one for situations where the new (and probably more complex) theory is necessary.

In your thought experiment, you said "relying on your ears only". This is as much a mistake as saying that time is measured on a wristwatch only, which is, of course, far too imprecise to test the effects of relativity at attainable speeds. As technology improves, so does our ability to measure and observe, giving us more data on which to base our hypotheses, such as the few-cycle timing differences in the playback. There may be just such a slight defect in our universe, that would open up a whole new field of study in trans-universal physics. There also may not be. We haven't observed such a thing yet, but we should not be so arrogant as to assume that our current knowledge is absolutely all-encompassing.

Re:Definitions

[Anonymous Coward]

Sorry we didn't get you before you hit puberty, but you would NOT want out of the fantasy at this point, believe me.

-Neo

Re:Definitions

[Anonymous Coward]

You have an mp3 player with two songs in it. The random playing algorithm makes it play the first song, the second song, the first again the second again and so on, because when it has to choose the next song there is only the other one available.

Your random number generator is at fault there. If it were truly random, it wouldn't have any qualms about playing the same song ten times before it hit the second song. If your random number generator is broken (and yours certainly is), it isn't random.

Re:Definitions

[Sarten-X]

Yes, everything I said was speculative. No, I don't think we're actually in a simulation. That does not mean that such a thing has been proven impossible. Until it is, it cannot be dismissed as a theory to explain observations that don't fit existing theories.

Science only deals in what we can see and observe

That is not limited to what we can see or observe right now. Perhaps a future observation will reveal discrepancies in We could not observe the Higgs boson until long after its existence was theorized. In fact, I think your original post said it well:

But given that we have no evidence to suggest we are, any assumptions around the notion that we are (or may be) are pretty much useless to us unless we can figure out the gaps in the simulation.

But where do we look for those gaps? What experiments should we run to determine whether that speculative instant communication is possible? We must consider the ramifications of such speculation, find a testable observation, then try it. That is curiosity. That is what drives discovery.

Re:Definitions

[Anonymous Coward]

That's what the people running the simulation *want* us to think!

Re:Definitions

[mestar]

It's promising in the first 25 minutes but then he wedges in "remote viewing" as a fact, and then it all falls apart.

Nothing later explains the world better, no predictions from his model where he puts consciousness into single atoms and similar bullshit.

Re:Definitions

[ZombieBraintrust]

Yeah from the article it seems they worked things out by writing a pet definition. There no experiment here. No observation. No math even. This is physics that could be done by lawyers.

Re:Definitions

[dargaud]

Yes, if we were in a simulation, we'd likely never know.

I'm not so sure. When you write a simulator, you use shortcuts in order to boost efficiency, speed, computing power, space, energy use, code reuse, etc... For instance you'd use jpegs instead of bitmaps for images. something inside such a simulation might notice that there are artifacts on images and deduce that something is not right. Another example is the use of procedural generation of objects (for instance trees in videogames): they are all different but all generated from a simple fractal+random algorithm. Of course how can you tell the difference between this and generation from (simulated) DNA+random events is left as an exercise to the reader.

Re:Definitions

[gstoddart]

What experiments should we run to determine whether that speculative instant communication is possible?

You don't. You study what you can know, and try to work out the details about it, and see if you can figure out something you didn't already know.

If you set out to solve a problem you don't have an inkling about how to solve (or any reason to believe it is solvable), you likely don't know where to begin and will just try random stuff.

We must consider the ramifications of such speculation, find a testable observation, then try it.

Well, you can enumerate a nearly infinite list of implausible speculations for which there's no evidence to suggest it's real. And you can design endless experiments which will tell you either your speculation is wrong, or that it's impossible to know (likely the latter in most cases).

See, unless you have something based on observations that says "hey, if I exploit this I can achieve something cool", you have a goal you'd like to end up with, and no idea of how to get there.

That's not science, it's wishful thinking.

So, if I say the universe is actually a hemorhroid on a giant space baboon's ass ... well, that's just wild speculation with no evidence to support it. And there is zero we can do experimentally to test it. Which makes is scientifically meaningless since it can neither be proven nor disproven.

Similarly, if I want FTL travel, instant communication or anything like that ... I need to have some evidence the laws of physics might actually allow for it. Barring any evidence to suggest it, it's not science.

Science doesn't make up an unsubstantiated theory and then test it. Science find evidence for something/observes something, develops a hypothesis, and tests it. You're skipping step 1 where you have something to suggest what you're looking for is real.

So, no, we don't always need to consider the ramifications of unfounded speculation as having scientific merit. That's essentially Intelligent Design.

That doesn't mean someone can't come along and change the rules about how we see things -- certainly Einstein did that. And that can radically change what you can do with what you know. But you don't start with an outcome you'd like to see and start trying to test it, especially when that outcome defies all known understandings of physics.

If there is a way to do instant communication over astronomically vast differences, or go faster than the speed of light -- we'd need to discover something else first, which would lead to the application.

Re:Definitions

[werepants]

Yes, if we were in a simulation, we'd likely never know.

But given that we have no evidence to suggest we are, any assumptions around the notion that we are (or may be) are pretty much useless to us unless we can figure out the gaps in the simulation.

The discrete nature of quantum mechanics certainly has some interesting implications and a lot of experiments have revealed that the universe behaves in a manner that seems almost like it has been optimized to run efficiently. The Stern-Gerlach apparatus, for instance, which measures spin of an electron, will only ever reveal it to be up, or down. Never in between. It should have an infinite variety of states, much like any physical object could be be right-side-up, upside-down, or anywhere in between. This is weird, and all quantum mechanical systems (and therefore reality) display this discrete behavior on some level.

Also, the double-slit experiment shows us particles that obey wave interference rules (kind of weird) even when there is no other wave/particle to interfere with (super weird).

Last, Bell's theorem (and the EPR paper) call into question whether or not things "exist" when they are not being measured. They apparently don't remember their properties in between measurements, and you would probably agree that it is reasonable to question the existence of something without defined properties. We actually have statistical analysis that proves that those persistent properties (called hidden variables) don't exist. That is what Schrodinger's cat is actually about - attach a cat-killing apparatus to one of these indeterminate particles and you end up with an indeterminate cat. The absurdity of that was the whole point of Schrodinger's thought experiment. It certainly would be a wise move to make sure that your universe didn't waste computational resources on parts of reality that weren't interacting.

Anyhow, all of these things show that reality is screwed up, and it doesn't take that much imagination to see quantum mechanics as one huge "gap in the simulation". Not nearly enough evidence to "prove" that we are in a simulation, but probably enough evidence to make a bit of philosophizing forgivable.

Calculating Space

[Anonymous Coward]

The idea that the universe computes comes originally from Konrad Zuse in 1967, who built the first working computer in 1941.

English translation of the Calculating Space book by MIT, 1969 (pdf)

Original article in German, 1967 (pdf)

You will rue the day when you discriminated

[Anonymous Coward]

against slime mold! The Great Ooze does not know mercy!

Different types of computation

[naasking]

The type of computation discussed in this article is not the type of computation used in the phrase "every physical event is a computation". These physicists are trying to discern computation from physical processes by discerning whether the process can encode information in its initial conditions, and other information can be extracted from its results. This is good when trying to determine which processes lend themselves to building computers, but it does not address the question of whether the universe is a computer, and whether the laws of physics are merely closed form equations describing some of its operational semantics.

Re:Different types of computation

[minstrelmike]

imo, the entire discussion and investigation comes down to one question: How do you define information?
IF DNA is encoded information and keeping a similar DNA 'alive' for billions of years is considered success (a Species is kind of a generalized unknowable DNA), then slime mold is successful. The randomization of sexual reproduction where all the stuff gets shuffled all the time (between generations), is one method of 'predicting' the randomness of the future. Or atl elast dealing with enough to get a good chance to try a second, then third, then fourth prediciton.

What exactly does being informed thru information mean?

Re:Different types of computation

[Anonymous Coward]

The type of computation discussed in this article is not the type of computation used in the phrase "every physical event is a computation". These physicists are trying to discern computation from physical processes by discerning whether the process can encode information in its initial conditions, and other information can be extracted from its results. This is good when trying to determine which processes lend themselves to building computers, but it does not address the question of whether the universe is a computer, and whether the laws of physics are merely closed form equations describing some of its operational semantics.

My knowledge may be dated here, but if--and this could be a big if--we accept the more main stream QM interpretations then there is honest-to-goodness randomness in the universe. That is, no hidden variables and no way of controlling & determining things like radioactive decay. This would imply that the universe is not a computer (or in my view, more than a computer), as it includes processes which cannot be predicted or computed.

useful/useless computation

[Anonymous Coward]

It seems to me that the differences expressed in the article is postulating may be better expressed by a distinction between useful and useless computation, which can than be defined as relative to the observer.
maybe that's just because I can't think of anything that we can observe, or infer, which has no effect on the physical world

I wouldn't necessarily call slime mold onecelled.

[Sique]

It's much more complicated than that. Myxogastria can be onecelled and mononuclear, and they can be multicelled and multinuclear, and they can even be onecelled and multinuclear - all within the same organism. A single plasmodium cell can contain up to 10 mio nuclei and span several square meters. Thus it would be better to call the plasmodium acellular, as it has no inner cell structure.

Adventures in epistemology

[Empiric]

For example, the processes that slime mould uses to solve a maze are largely unknown. For this reason it is not computation.

Don't we usually declare characteristics of things based on what we know about them, rather than on the basis of not knowing about them?

Seems like a strange kind of subjective solipsism--"what is, is dependent upon on what I currently know is".

Re:Adventures in epistemology

[nashv]

Yeah, who the hell are these authors anyway? Engineers?

Re:Adventures in epistemology

[Anonymous Coward]

You may be right. On the other hand, the slime mold, when questioned closely, admitted that it did not know how it solved the maze either, and thus was in no position to dispute the idea that it is not calculating. While not dispositive, it is certainly a major concession by the slime mold.

Re:Adventures in epistemology

[Hognoxious]

Don't we usually declare characteristics of things based on what we know about them, rather than on the basis of not knowing about them?

I'm Donald Rumsfeld, you insensitive clod!

Let's Test Some of These Assumptions

[Anonymous Coward]

"Without a good understanding of the underlying physics and solid engineering to make the device usable, a physical system simply evolves rather than computes." Okay - Is the Earth a computer? The earth, as a device, is definitely usable - it is specifically "engineered" to sustain human life, unlike any other system in the universe. We have a good understanding of it and we've engineered subsystems within it to even make it more usable. In fact, it generates spreadsheets (I have yet to see a quantum computer generate one) in some of its subsystems.

So, using the blog post's assumptions, the earth doesn't just evolve, it also computes. Correct?

Semantic triviality

[oldhack]

See subject.

Re:Semantic triviality

[Suiggy]

Agreed. The authors of the papers are just old biologists who don't like the idea of pancomputationalism and digital physics infringing upon their domain. They want the magic of life to remain elusive. The truth, however, is that computation *is* the magic they so desire.

BS

[nashv]

I RTFAed. Their theory is essentially that computation can only be said to have occurred if you know the physical nature / laws that allowed the computation to occur.

Which is BS. There are plenty of people who can add 2 numbers on a calculator without knowing anything about electrons, bits, electronics etc. You can extend that until the number of people who understand specific physical laws underlying a computation is zero.

Since when is human knowledge the test for whether any computation is happening? All they are saying is "If we don't understand it, we will not call it computation." Way to go with the semantic circus.

Re:BS

[NoNonAlphaCharsHere]

They're trying to draw a distinction between computation and experimentation. Essentially, they're saying that if your model/mental-map of the problem is good enough, i.e. you have good predictive power over your methods, then you are performing a calculation/computation. If you're not so certain of the outcome, then it's experimentation. But I do agree, it's kind of a semantic knife's edge.

Re:BS

[wonkey_monkey]

Since when is human knowledge the test for whether any computation is happening?

Are you sure anyone's referring to human knowledge? I've only skimmed it but got the impression that their criterion is that there is some information (in the new-fangled sense) to be had at the end of it - it doesn't have to interpreted by a human, it simply has to be available (to a human, or a cat, or a particle in the vicinity).

Re:BS

[nashv]

There is always information around (new-fangled sense). In fact, Any dissimilarity in the universe of any kind is information. Presumably, the ability to recognise such dissimilarities is required for any information retrieval, and in the end humans must detect it if we are to talk about if it is computation or not.

The point they are making is that one must understand a systems physical laws governing events for it to be a computation. Wrong. What they say actually applies to the fact that we must understand physical laws to use a sequence of physical events as a computer. In fact, the computers we use are simply devices where we can set up physical conditions in such a way that they reflect our problem (encode) and then let the physics solve it. That ability to set up physical events in particular ways makes encoding easier. But it is NOT a requirement. You could develop an encoding scheme that allows you to put in your problem in terms of whatever states a system has. That is exactly the principle behind DNA computing. The physical events of DNA polymerization and annealing of complementary strands are computing all the time. If the information makes any sense to you then, you have not magically invented computation, you just discovered a way to use a physical system as a computer for your pet problem.

All computation needs is that some physical events happen in a predictable manner (whether you understand why or not). I am doing the computation all the time, but nobody knows why positive charges and negative charges attract. It's not important to use electrons as a computer as long as I am sure they do. The same thing applies to their slime mould example. It solves a maze. It's. I don't know how it does it...but it predictably does. And that's a computation.

Re:BS

[Anonymous Coward]

A couple of generations ago, scientists regularly wrote animals were thought to have no feelings, no tools, and the most rudimentary communication skills, and homos suffered from mental illness. A couple of generations before that, most human beings who didn't subscribe to European conceits were savages.

Now scientists are entertaining ideas about the whole universe like, the whole thing began with an explosion, and we're so clever that we can model it, that all the mass in a star can shrink into a tiny dot, that 3/4 of the universe is made of particles we haven't seen in a century of looking, and that there is no such thing as consciousness.

Phlogiston lives and will always live.

Re:BS

[StripedCow]

And what is this thing they call "information"?

In a deterministic universe, if the universe starts with X amount of information, then the amount of information can only drop I suppose. Because determinism implies that a computation is going on, and computations can only destroy information, not create it. For example, performing x^2 destroys information, because we lose information about the sign of x. But performing 2*x keeps information, because we can always compute x back from 2*x.

So, I suppose that "information" can ONLY be created from the non-deterministic part of the unfolding of the universe. When there are at some point N non-deterministic choices, then this accounts for log2 N bits of information.

Is this the correct view?

Universe does compute - per se.

[Codifex Maximus]

I think whether or not the Universe computes is relative.

What is it you're trying to compute? If you are trying to compute the dynamics of the Universe then the Universe does compute. It computes itself better than any other model.

If you are talking about abstractions then probably computers do very well give proper instruction and dataset.

To get a computer to compute the universe is like trying to force a very large round peg into a very small square hole.

When in doubt, ask Einstein.

[briancox2]

"God does not care about our mathematical difficulties. He integrates empirically." --Albert Einstein

42

[sycodon]

Slashdot is slipping. Twenty Five posts and not one mention of Douglas Adams.

Re:42

[wonkey_monkey]

We're all waiting to claim post 42, of course.

Re:42

[dgatwood]

The mice were mentioned in the "department" line, so we assumed we'd be modded redundant. But as long as we're going down that path....

Ooh. I know this one. The computation stops when we get bulldozed to make room for an intergalactic superhighway.

Re:42

[DutchUncle]

I'm shocked, shocked and dismayed, that I had to scroll down this far to find that someone had beaten me to it.

The universe computes on third shift

[nani popoki]

On first shift, you submit the keypunch forms.

Slime mold does perform computation

[wiredlogic]

Slime mold senses its environment and reacts accordingly. That is computation in the broader sense used by physicists. It isn't engineered (unless you are of a certain ilk) but it is an organized, coherent system.

Toward Egan/Wolfram Territory

[Nova Express]

I.e., the idea of the entire universe as constituting some sort of universal computational substrate.

The idea is probably wrong, mainly because every "my conception of the fundamental nature of the universe based on just discovered science" is wrong, due to the time-bound nature of our perceptions.

Re:Toward Egan/Wolfram Territory

[presidenteloco]

>The idea is probably wrong, mainly because every "my conception of the fundamental nature of the universe based on just discovered science" is wrong...

probably wrong but not necessarily wrong.

Maybe, if people keep coming up with new kinds of science (pun intended), one new theory will finally get a lot of the rest of the unknown bit (pun not intended) right.

Just because computing is the "latest thing" does not mean it is not a better analogy/explanation of certain things (minds, big chunks of physics) than for example, the steamworks, or electrical circuits, or levers and wheels analogies from before. To think so would be falling into another "believing the conventional wisdom" trap.
Some theories are just better (or more comprehensively explanatory) than others. It doesn't really matter whether the theory is new.

Re:Toward Egan/Wolfram Territory

[Anonymous Coward]

The idea that the universe computes comes originally from Konrad Zuse in 1967, who built the first working computer in 1941.

English translation (by MIT) of the Calculating Space book, 1969 (pdf)

Original Rechnender Raum article in German, 1967 (pdf)

Proof by axiom

[Froggie]

By defining computation to be X and not Y, I have proved that X is computation and Y is not.

Yes, terribly helpful, that.

Planck physics

[Anonymous Coward]

It's probably wrong, but I really like the idea of the universe being a big physics engine with a step time of one planck second.

The computation is ruined

[cold fjord]

Based on the story one might speculate the all processes in the universe could produce a computation and provide an answer. If so, then the universe has been calculating in this story as well. The number of posts in this story prior to this post is 42. My post now invalidates the universal answer by changing the number of posts to 43.

Angles and Pinheads

[pseudorand]

Does this remind anyone else of how religious philosophers of days past used to argue over how many angles could dance on the head of a pin? I'm not sure about the angles part, but there are surely some pinheads in this story.

Re:Angles and Pinheads

[MachDelta]

Let me guess - those dancing angles are doing the tan-go?

Re:Angles and Pinheads

[Anonymous Coward]

Do you know the difference between angel and angle?

Re:Angles and Pinheads

[pseudorand]

Apparently not, since I did it multiple times. I blame the US public education system.

Re:Angles and Pinheads

[Anonymous Coward]

non Anguli sed Angeli.

Re:Angles and Pinheads

[hajus]

They were debating the concepts of infinity.

Re:Angles and Pinheads

[Anonymous Coward]

I was definitely reminded of that.

Wrong.

[VortexCortex]

However, that may need to change now that physicists have worked out a formal way of distinguishing between systems that compute and those that don't. One key is the ability to encode and decode information. 'Without the encode and decode steps, there is no computation; there is simply a physical system undergoing evolution,' they say. ... So slime mold fails the test while most forms of quantum computation pass."

The slime mold is computing the optimal solution to the task at hand. Evolution is the computation, the DNA contains the encoding and decoding. DNA is self describing information -- A proposed solution to the problem space of the environment. I just despise when morons fail basic Cybernetics 101. Explain reproduction if there is no encode / decode phase... ugh.

Well Duh...

[Anonymous Coward]

... at night of course...

Alternate Universe

[Anonymous Coward]

+++ Divide By Cucumber Error. Please Reinstall Universe And Reboot +++

Etc.

[Impy the Impiuos Imp]

Physicists, please leave this to computer people. Interpretation is just a transformation, and is no magic door to computational efficiency.

If a slime mold could, via interpretation, solve some NP-Hard problem, that would be an astounding result with major implications for the computational ability of the universe.

This is independent of the interpretation. It is the equivalency of reducing a problem to another, of which physicists are indeed well aware.

When does the universe compute?

[Anonymous Coward]

I think the phrase "spread like wildfire through science" is a little over the top.

As I recall, Edward Fredkin's interesting ideas about Digital Philosophy (roughly speaking, the universe as a computer), are far from being accepted by the scientific community.

Nethack

[CanHasDIY]

This slime mold is delicious!

Lets rock this topic like its the 90s!

[YoungManKlaus]

after all that's roughly how current it is ... I remember learning about these topics in bio (slime mold) and physics (quantum computing) back in high school (or rather the equivalent in my country), and the material was already some years old - and already claiming "10 to 15 years" ... well, that timeframe was over last year iirc

computation only comes from human programming

[globaljustin]

this is a true statement:

Why can we not think of the information as being embodied in "some aspect or other of" the matter and energy undergoing evolution. It is only some observer that needs to see the information as having been encoded or decoded.

I think TFA (and Turing too, but we'll get to that) make a fundamental mistake that leads them to these TED-talk kind of wild eyed pop-science speculation.

It relates to social construction of reality: http://en.wikipedia.org/wiki/Social_constructionism

So, my point is: the very *idea* and *definition* of the term 'computation' is a human-created contextualization of the way we observe the universe to work.

humans made machines (computers in this case) to automate a task: specifically in the case of the 1st computer it was for automatically doing long math problems.

all of 'computation' in the history of the known universe has been done by machines programmed by humans

that's what it is

the deeper problem, IMHO, is a computing model based on the 'Church-Turing' Thesis of universal computability to describe computation...its sort of like an unnecessary abstraction layer

it's time for a better model...information theory and communication theory are helpful here...I believe a *cybernetic* model is the proper understanding of computation.

Why? Computation is nothing more than humans observing information flow through a system, a system which *processes and changes* the information in a predicted and predetermined (*programmed*) way.

For this, I see the Shannon-Weaver Communication model as the proper conceptual starting point for formal understanding of 'computation'

Just imagine Von-Nueman Architechure overlayed the Shanon-Weaver Model diagram...**that's** a cybernetic approach ;)

Re:computation only comes from human programming

[retchdog]

information theory tells you the limits of what you can know, without regard for how hard it might be to get there. (actually it mostly proves upper bounds on these limits, but that's the style of it.)

computation theory speaks of how hard it is to get there.

if your opponents have only a classical computer and polynomial time, then you don't actually need a one-time pad (which is an information theoretic limitation); you can fake it, and they won't have enough time to unweave your trick. in the other direction, if i know a message X, and give you a one-way invertible function applied to X, i haven't really told you anything, except that I've given you a way to check if someone else (under appropriate assumed restrictions) probably has exactly the same message; you can request they apply the same function and give the result.

the thing is, roughly speaking, Shannon theory depends on a probabilistic model (or something very close to it), while computation theory does not. computation theory can analyze probabilistic computation, but it's optional. i'd like to see a derivation of an equivalent of (or refinement of!) P and NP and NP-completeness from information theory. i don't think it's possible, and those are pretty important ideas in computational complexity.

Sounds like nonsense to me

[gweihir]

"A physical system undergoing evaluation" is pretty much what a computation on a physical device looks like. Seems some physicist(s) with their usual search for meaning when the midlife-crisis strikes but little idea of Computer Science drummed something up that does not make a lot of sense. The issue is of course that the encoding and decoding steps (properly called abstraction and application) are only necessary when you have a computing device using a different primary mechanism that the device the calculation apply to. For example, when doing analog computations in a radar system, you do not change the mechanism, the signal already is electrical and analog. Hence when doing, day, computations for gravity, there would well be an invisible tiny "gravitational computer" in any particle affected by gravitation that uses the available input directly.

Of course, there is always the (likely) possibility that this whole universe is only a simulation, and all perception of it we have is a cleverly crafted illusion. In that case the abstraction step is not necessary either, except possible in the bi-directional channel that delivers the illusion to each of us. (Solipsists win the most here: Only one bi-directional channel needed for the whole universe ;-)

Re:Sounds like nonsense to me

[Anonymous Coward]

Of course, there is always the (likely) possibility that this whole universe is only a simulation, and all perception of it we have is a cleverly crafted illusion. In that case the abstraction step is not necessary either, except possible in the bi-directional channel that delivers the illusion to each of us. (Solipsists win the most here: Only one bi-directional channel needed for the whole universe ;-)

You seem to be assuming that if it were a simulation then we would have to exist in some capacity outside of it, like in the Matrix. Why is that? If it is just a computer running a simulation based on the laws of physics then we could be just bits of memory in the computer running the simulation. But of course from our perspective we would still be "real".

Re:Sounds like nonsense to me

[Anonymous Coward]

It doesn't matter whether universe is a simulation or not. What matter is that we are aware of the possibility. Are dolphins aware? Chimps? You explain chimp what it means to nuke something from orbit. I will be impressed.

Re:Sounds like nonsense to me

[gweihir]

I am a dualist. Physicalism has some very obvious problems, just as theism has. So I assume the simulation would only provide the "reality interface", and not create the individuals in there.

Re:Sounds like nonsense to me

[gweihir]

While I do not get your animal references at all, I agree that awareness of the possibility is key.

Cue the dominant paradigm...

[hawkfish]

In the 18th century, the sexy model of the universe was a clockwork mechanism because that was the coolest tech available at the time. Move forward three centuries and suddenly the universe is a digital computer because that is the new trendy tech. Never mind that 50 years ago Feynman showed that the universe is not efficiently computable by a Turing machine...

Re:Cue the dominant paradigm...

[nashv]

This is what Feynman said

""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 space/time is going to do? So I have often made the hypotheses 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".

. There is a subtle difference.

Re:Cue the dominant paradigm...

[Anonymous Coward]

Feynman also hypothesized that a quantum computer CAN efficiently compute the universe. Seth Lloyd later proved that and he suggested (in his book "Programming the Universe") that the universe is indistinguishable from a quantum computer computing the laws of physics. The fact that a classical computer cannot do it efficiently is irrelevant.

Re:Cue the dominant paradigm...

[iggymanz]

that's probably a rather silly statement, what with a true quantum computer and the universe apparently employing quantum effects and all.....

no, not really

[stenvar]

However, that may need to change now that physicists have worked out a formal way of distinguishing between systems that compute and those that don't.

Just because a bunch of physicists propose such a "formal way" doesn't mean that that formal way is actually correct or even meaningful. The history of science is littered with physicists proposing all sorts of things outside physics, almost all of which have turned out to be b.s.

Sounds like a love child of physics and philosophy

[Anonymous Coward]

Philosophy always just makes my head hurt beyond a certain simple level. At least physics makes my head hurt for a purpose. I'd like to get something out of my headaches other than a paper that only others of my kind will read. At least... that's my philosophy...

Searching

[b4upoo]

Philosophers tend to want to create specialized definitions and terms in order to project that philosophy so that it can be more broadly applied. I think this is what we are seeing here. In specific they mentioned slime mold and encoding and decoding. Actually slime mold is encoded with DNA and the function of the slime mold in navigating the maze reveals the decoding expressed as the function of slime mold. If our human abilities were more advanced we should be able to determine the future functional abilities or probabilities of a function by examining the DNA encoding directly rather than determining its function by observation of the slime molds activity.
                    Essentially they seek clarity on what defines computation. In this muddy world clarity is rarely available.

Obligatory SMBC

[Anonymous Coward]

As usual either Zach Weinersmith or Randall Munroe, the two greatest minds of our era have already forsaw this news...

In this case it's Zach!
http://www.smbc-comics.com/?id=3054

Pen and Paper?

[pscottdv]

Computation is essentially the same process. A computation uses the evolution of a physical system to model an abstract theory.
But this only works when the link between the real and abstract worlds is clear and well understood.

Yet strangely I can work computations using pen and paper using only the evolution of the poorly understood system consisting of my brain, arm, hand and eyes.

Re: I don't think encoding/decoding are fundamenta

[Anonymous Coward]

Virtual particles and the probabilities of quantum particles to pop into existence? Haven't we seen an article on /. Proving the measurable existence of these two phenomena?

I think that is what the prior poster was referring to.

When all you have is a hammer...

[govett]

Scientists quantify the universe, so hither, thither, and yon, all they see is computations.

Re: I don't think encoding/decoding are fundament

[Anonymous Coward]

Or are we not sure if the laws of thermodynamics in quantum systems is actually inaccurate because we are not sure the universe is a closed system.

Living in the Matrix

[quax]

Maybe that'll throw a monkeywrench into the viral notion that we are living in a simulated universe.

Not holding my breath though.

Spacetime..

[Anonymous Coward]

Some call it "general relativity," but I call it network lag.

what someone else said

[globaljustin]

hey thanks for the response...it's clear you are more adept at handling the terminology of Computer Science.

IMHO, though you are batting around heady ideas, it is all just language to describe **things we program machines to do**

i'm not being reductive_it's recontextualizing fresh language to describe what we observe and, in that context **what we can predict and how**

start with your introductory sentence...'X theory does 1, while Y theory does 2'...wrong...humans test hypothesis and program machines to execute instructions

information theory tells you the limits of what you can know.....computation theory speaks of how hard it is to get there.

this isn't something I can argue for or against...I won't have a definition war, but its safe to say these arent formal definitions in any way

that's what's happening...the language of 'computatility functions' is useful no doubt, but conceptually it is a Moebius Strip of logic

to be specific:

if your opponents have only a classical computer and polynomial time, then you don't actually need a one-time pad (which is an information theoretic limitation); you can fake it, and they won't have enough time to unweave your trick.

and

the thing is, roughly speaking, Shannon theory depends on a probabilistic model (or something very close to it), while computation theory does not.

Wrong on both counts.

1. This isnt a crypto-contest...that's part of your problem, computing is programming a machine to follow instructions...you are describing a cryptographic scenario that is ***ANALOGOUS*** to the same problems found in computing....but it is only an analogy

2. The Shannon-Weaver model originally was devised to reduce 'signal noise' in wireline transmissions...it is *not* inherently related to probability. Probability is a way we use math to observe and calculate the 'noise'...IT IS ONLY METRICS AND MEASURES

Look at the scientific definition of the speed of light...it's a human attempt to quantify somethign we observe in nature...we define light's speed based on its relationship with other known aspects of the universe.

You are confusing the act of defining the 'noise/randomness/etc' of a signal quantitatively with theory about how the system of communication/computation works.

I know that alot of CS types have a whole schpiel about 'computational complexity' and 'p/np' stuff, but it is not what you think it is, from a functional perspective.

The entire notion of the Turing thesis is off-kilter and not truly a theory of computation. It is a **thought experiment** about computation. That is ALL.

tl;dr We can compute anything we can program a machine to compute...that's the limit of computation.

Re:what someone else said

[retchdog]

i'm sorry, but without definitions we're not going to get anywhere. it's been a while since my formal training in complexity theory, but i don't think i've committed any fundamental errors; i'm trying my best to both recall and summarize for a public forum, while you do nothing but bat around baseless assertions and try to insult me. there's nothing wrong with baseless assertions, but please don't assign them labels which you don't understand. consider refining your ideas into precise statements, rather than disparaging this as some kind of triviality you've transcended.

particularly, if you'd just follow the links from the wikipedia page you cited, you'd see the importance of probability in the formulation of Shannon information.

so, entropy, in other words...

[globaljustin]

Now, discounting Entropy doesn't mean I deny that systems have a tendency to reach a Steady State where a perceived equilibrium has been established.

you just described the heat death of the universe

which is another way of saying 'entropy'

therefore, you actually 'adhere' to the Theory of Entropy...at least as you define it.

that's the problem, the concept of 'entropy' has been wrapped inside itself so now it resembles a Moebius Strip not a defined term for somethign we observe in the universe

THE ACT OF DEFINING 'STEADY STATE' , as you call it, is the same thing as defining 'entropy' ...its just using different language.

I don't know where/when/how 'entropy' became a politicized word in science, undoubtedly some religious jerk tried to say that b/c of what Newtonian Physics tells us (its not a completel theory but it is useful for prediction) and b/c the universe ends in heat death, then therefore God exists...that's my guess

Re:so, entropy, in other words...

[Anonymous Coward]

Does steady-state actually exist in the universe, or is that just an assumption?

this^

[globaljustin]

gotta agree...

IMHO it comes down to inherent faults with the Chuch-Turing 'thesis' and computational complexity theories

question of language

[globaljustin]

but it does not address the question of whether the universe is a computer

b/c, IMHO, 'computing' is the act of programming a machine with instructions

that's all 'computing' is...asking if the universe is a 'computer' is like asking if it is a 'Volksvagen'....they are both just analogies of types of systems

whether the laws of physics are merely closed form equations describing some of its operational semantics.

as opposed to what?

how could 'the laws of physics' be anything other than humans attempt to make heuristics that can predict nature

Re:question of language

[cowdung]

Computing is not the act of programming.. Computing is the act of executing a program.

In my opinion, this is a question for Computer Scientists and not Physicists.

Re:question of language

[naasking]

b/c, IMHO, 'computing' is the act of programming a machine with instructions

Computing and programming are not synonymous. Whether a machine is a programmable computer is a completely separate question from whether it's a computer at all, ie. the universal Turing machine vs Turing machine distinction.

as opposed to what?

Reality could compute uncomputable functions (like functions on the reals). That would make natural laws super-Turing, ie. Turing + Oracle. The Oracle isn't a computer as we currently define it.

When does the universe matrix?

[AdamThor]

"The idea that every physical event is a computation has spread like wildfire through science."

Not that it's necessarily true, but I think it would have made better science fiction had The Matrix claimed that humans were computation units in the Machine's world, rather than that "battery" stuff.

[Morpheus voice]
What an amusing reversal... Which lane is fastest through this interchange? How many people get in the elevator? Have the soup or the salad? Tastes great or less filling? Our surroundings are encoded problems, and the aggregate of human behavior presents a solution to those problems. How many people make a quantum bit? 1,000? 10,000? 1,000,000? It depends perhaps on the resolution of the solution that is sought. What questions are we answering for the machines? Who could possibly know?
[/Morpheus voice]

They could have explored that, maybe, in a sequel, had they ever made one. Shame they didn't get around to it...

Douglas Adams predicted this

[Anonymous Coward]

One hopes no one demolishes the Earth before the computation is done though.

The Universe...

[mug funky]

does not compute.

Seriously, this has to be the worse MMORPG ever.

[Anonymous Coward]

"SHUTDOWN"

Who the hell comes up with games like this?

Compiler

[relisher]

What happens when the universe has compiler errors?

**still** a question of language

[globaljustin]

Computing is not the act of programming.. Computing is the act of executing a program.

programming is not computing?

when I write a python script, that isn't under the umbrella of human activity we use the word 'computing' for?

you're dead in the water on this one...I'd like to have a discussion about this topic but this is not that

Re:**still** a question of language

[bingoUV]

You are the programmer, and the computer is the programmee. When you write a python script,

1. You, the programmer, is programming.
2. No one is computing (ok, your IDE is computing, but no computing is happening as per YOUR python script is concerned).

When you EXECUTE your python script,
1. The computer is computing.
2. No one is programming.

My!

[ihaveamo]

That was a yummy SLIME MOLD!

I know Nethack was an old game, but I didn't know it was written with organics...

X is Turing b/c Turing things are X

[globaljustin]

Thanks for the response.

I think my whole point is that Turing's concepts should be done away with...so when you say this:

Whether a machine is a programmable computer is a completely separate question from whether it's a computer at all, ie. the universal Turing machine vs Turing machine distinction.

I want to scream (the truth)...the discussion itself (like the one we're having) begins to resemble a meta-Turing Machine itself...then I feel crazy inside....

Turing is as Turing does. Anything can be 'Turing-complete' or w/e term you use b/c the whole notion is just a thought experiment about computing that created a false dichotomy ('turing complete or not') that *becomes* a testable question only b/c that linguistic distinction (not a real distinction, just an analogy carried through).

I think, unfortunately, we may have to settle for 'agreeing to disagree'...I really want to convince you of what I'm saying, or at least really get through to you and hear your reaction.

Here is is: http://slashdot.org/comments.pl?sid=4315105&cid=45073983

I spelled it out much better on a different part of the thread for this story...I *promise* you'll find a coherent, interesting, and hopefully convincing argument for all of Computer Science to move away from Turing model of computing.

I think it's time to be done with Turing.

Re:X is Turing b/c Turing things are X

[naasking]

Anything can be 'Turing-complete' or w/e term you use

Anything cannot be Turing complete. There are whole classes of useful languages that aren't Turing complete, like regular expressions, cellular automota, and total programming languages. Turing complete is a useful distinction because it tells you whether you can check various incompleteness properties of programs written in your language, like checking whether your program terminates.

And Turing machines aren't the only way this has been approached. Church's lambda calculus is even more widely used in computation literature. Combinator calculi are also sometimes used. But there is a formal equivalence between them all, so the choice of one immediately implies results for the others.

Turing completeness is not going away, and it's not harmful to CS in any way.

Re:Saxons and Boneheads

[Hognoxious]

Non Angli sed Angeli.

-- Pope Gregory I

Not much to see here, move along

[Doghouse13]

This isn't a discovery or a proof of anything. At heart, it's simply a (deeply academic) attempt to usefully define "computation". From the abstract: "In this paper we introduce a formal framework that can be used to determine whether or not a physical system is performing a computation." In other words - "We've developed a definition of computation, which we think is useful, under which some physical systems turn out to be performing computation, and others don't." If you subscribe to the view that all physical events involve computation, that's not an ideal way of putting things - but, even if you don't subscribe to it as a definition of what is or is not computation per se, it still provides a way of classifying physical systems, that may be of use.

facial tissue vs 'Kleenex'

[globaljustin]

a checking method is not fit to be a fundamental theory...its building a castle on sand

i will agree that *in practice* Cambridge has successfully claimed the concept of 'computatiblity' with its 'Turing' brand

That's why i'm calling for a better theory

in the context you describe, 'Turing-completeness' is nothing more than a brand name like Kleenex or Charmin

Turing complete is a useful distinction because it tells you whether you can check various incompleteness properties of programs written in your language, like checking whether your program terminates.

ok...I understand where you are coming from, but again, it's just a **linguistic** or **jargon-based** distinction...

checking to see if a program terminates can be done several ways...as you yourself admit:

And Turing machines aren't the only way this has been approached.

exactly...checking a program isn't exclusive to 'Turing'...it's a name brand of an essential function of programming

analogy: changing the oil in your car

there are simple steps that anyone who understands the basics of cars would understand

calling the act of 'checking your program to see if it terminates' a specific theory is like saying that you have the Nassking Method of changing the oil which requires filling the car up with fresh oil after the old oil is drained.

you could write all kinds of papers why this is important, devise tests to prove it is important, but the fact is, you're using jargon to *brand* an essential activity

programs need to be check for 'completeness' & or 'proper functioning'

as a 'fundamental' theory to somethign like Computer Science the 'Church-Turing Thesis' is not fit...it is a methodology based on an analogy of computation

computation is machines executing human-written instructions

interesting

[globaljustin]

makes sense as a dichotomy in that context...

nested above someone mentioned the 'question' of whether the 'universe is a computer' which started this mess

i can't quibble with your contextualization of the terms, and using your definitions, I'd have to say that the question of whether the 'universe is a computer' is functionally the same as asking if the 'universe has a programmer'

*that* is an interesting question...one which science isn't fit to answer

what i mean is, the question of the existence of 'god' relates and involves science, but by definition isn't answerable scientifically

we can't prove or disprove the existence of 'god'

that's IMHO