Uncommon Descent Serving The Intelligent Design Community

Landmark 1929 Physics Paper: On the Decrease of Entropy in a Thermodynamic System by the Intervention of Intelligent Beings

Share
Facebook
Twitter
LinkedIn
Flipboard
Print
Email

In the Forgotten Creationist/ID book endorsed by Nobel Prize Winner in Physics, there was mention of a landmark physics paper by Leo Szilard. The paper dealt with a long standing problem of Maxwell’s demon. The paper was not immediately about the problem of design, but it sparked later questions by Robert Gange and others about whether a mindless biotic soup can somehow have sufficient insight about itself so it can build the computerized self-replicators that characterize life.

First, Gange describes the problem of Maxwell’s demon, and why Leo Szilard’s solution was important:

Maxwell’s Mysterious “Demon”

The organizational intricacies of protein reflect information on a scale that a Supreme Intelligence can produce, but that nature cannot. To see why this is true, let’s think about a small imp who became known as “Maxwell’s Demon.”3 We will allow the imp to control a tiny window that connects two adjoining compartments. In your mind’s eye, imagine two boxes joined by a common wall. In the middle of the wall, picture a tiny window that connects one box to the other. On one side of the window there’s a shelf where the imp is perched.

The imp is able to open or close the window at will, and without effort. Both boxes contain air and from time to time, as a result of this air, a gentle breeze blows against either side of the window. The imp is told, “Open the window if the breeze on your side is strong; otherwise keep it shut.”

Now this may seem like a simple request, but the question is, can the imp obey the instruction? Although it may seem like something he can do, it turns out that were he to successfully perform the required task, he would violate one of the most fundamental laws of science. It’s worthwhile to learn why this is so, because we will not only uncover a fascinating insight regarding the origin of life, but we will also discover the answer to something that stumped the whole world for over half a century regarding Maxwell’s demon.

We’ve said that each of the two boxes contains air. But air consists of tiny molecules which are atomic specks so small that about 10 thousand billion will fit onto the head of a pin. Furthermore, these miniscule dots are in a state of constant motion; we sense them each time we feel a breeze. A strong breeze means that they’re moving fast whereas no breeze means that they’re hardly moving.

Now suppose the imp opens the window each time a strong breeze occurs. If he consistently does this, all the fast-moving molecules will pass through the open window and into the box on the other side. But since he keeps the window closed when there is no breeze, the slow molecules will remain in the box where he’s standing. Thus, the imp has succeeded in separating the fast and slow air molecules, putting the fast ones into the one box, and keeping the slow ones in the adjoining box. From a scientific point of view, faster air molecules mean a higher temperature and an increase in pressure.

Therefore, our imp has created a pressure and temperature difference between the two boxes; i.e., he has created energy!

But how can Maxwell’s demon work? How can he create energy? This question baffled the world for many years, and no one was able to offer a satisfactory answer. Scientists asked, “Why can’t an imp open a window?” If he could, he certainly could separate the fast- and slow-moving air into separate compartments. The fast-moving molecules will travel to one side, and the slower ones will remain in the other.

No one questioned the fact that, at least in principle, the imp had created energy. Let’s see how we know that this is true. To show that the imp has created energy, we can wait until he’s collected all the fast-moving air on one side. When that is done, we’ll open the window, but this time keep it open. Air from the high pressure side (box with the fast-moving molecules) will rush through the opening and into the other side. If a generator wheel is located near the window during the time it’s open, the resulting gush of air can be made to turn the wheel of the generator and, thereby, make electricity. Therefore, the imp does create energy. But here’s our dilemma: It’s impossible to create energy in a closed box! So no one could figure out how the imp could do it!

Death of a Demon

Maxwell invented his demon in the 1800s, but not until 1929 did a scientist named Leo Szilard find the answer. The imp can’t create energy — not because he’s unable to open and close the window, but because he doesn’t know when to do so. In other words, he doesn’t have the information necessary to identify which air molecule is moving fast and which is moving slow. But what’s even more important, it costs him energy to acquire the information he needs! In fact, Szilard did a careful analysis showing that it costs more energy than the imp can recover.4 Simply put, the process of creating energy forces you to lose it! We can phrase it yet another way: Information is equivalent to energy in the sense that to have one means you can create the other.

Gange extrapolated the ideas in Szilard paper and argued that something in a disorganized state like a biotic soup cannot self-organize itself because the problem of a soup gathering the requisite information to build life in a closed system will actually create more disorganization! Hence life will not spontaneously arise in a closed system.

Gange asserted that the problem of a system spontaneously creating novel life-giving information is like the problem of perpetual machines. Whether Gange was successful in reframing Szilard argument for perpetual motions machines into the problem of OOL, I’ll let the reader decide, but here is the landmark paper by Szilard that was finally available in English:

On the Decrease of entropy in a Thermodynamic System by the Intervention of Intelligent Beings

UPDATE:
I just found some nice info on Leo Szilard from wiki. He was a first rate scientist. From wiki:

Leó Szilárd (Hungarian: Szilárd Leó; German: Leo Spitz until age 2; February 11, 1898 – May 30, 1964) was a Hungarian-American physicist and inventor. He conceived the nuclear chain reaction in 1933, patented the idea of a nuclear reactor with Enrico Fermi, and in late 1939 wrote the letter for Albert Einstein’s signature that resulted in the Manhattan Project that built the atomic bomb.[1] He also conceived the electron microscope, the linear accelerator (1928, not knowing Gustav Ising’s 1924 journal article and Rolf Widerøe’s operational device) and the cyclotron.[2] Szilárd himself did not build all of these devices, or publish these ideas in scientific journals, and so credit for them often went to others. As a result, Szilárd never received the Nobel Prize, but others were awarded the Prize as a result of their work on two of his inventions.

The photos above is Szilard with Einstein.

PS

I think the ID community has very very good arguments against OOL outside of thermodynamics. I’m hopeful that the study of thermodynamics will help the ID cause, but it must be an informed and careful study. I provided the above paper for those interested in the technical details of the subject.

Szilard dealt with thermal entropy, an interesting project for the ID community is reframing Szilard argument into an ID argument and in a way that is accessible for popular consumption. Right now some of the arguments are a bit too technical and need a little more rigor.

Comments
[Gange] says that the demon supposedly creates energy, but in fact (if it worked) it would decrease entropy (thus making the energy that was already there avalialble to perform work).
Agreed. It was a misstatement. Even though this was for popular consumption it was poor choice of words. I would not however rush to judgement that he doesn’t understand thermodynamics, his writings don’t seem to reflect that.
Also agreed. When I've gone through Gange's book properly I'll be much better able to judge it on its own merits.
...but more recent analysis...
The book was written almost 30 years ago, and Szilard’s paper 84 years ago.
Good point, I wasn't taking that into account. As far as I know the "modern" understanding dates from a 1981 paper by Charles H. Bennett (athough the seeds were sown by Rolf Landauer in 1961). Gange is essentially following Léon Brillouin's extension of Szilard's understanding, which was quite widely known in the physics community (even after Landauer & Bennett showed it was basically wrong).Gordon Davisson
December 12, 2013
December
12
Dec
12
12
2013
11:55 PM
11
11
55
PM
PDT
Nice to see you Gordon, I appreciate you weighing in.
Gange also misstates both the problem of Maxwell’s Demon and the solution in significant ways: he says that the demon supposedly creates energy, but in fact (if it worked) it would decrease entropy (thus making the energy that was already there avalialble to perform work).
Agreed. It was a misstatement. Even though this was for popular consumption it was poor choice of words. I would not however rush to judgement that he doesn't understand thermodynamics, his writings don't seem to reflect that.
). He also says that the resolution of the demon paradox is that “it costs him energy to acquire the information he needs”, but more recent analysis has shown that information can be gained “for free”;
The book was written almost 30 years ago, and Szilard's paper 84 years ago. To argue that doesn't bode well for Gange's understanding is quite a stretch especially since the question of Maxwell's demon is not a primary concern for most operational thermodynamics. The original conception of the Demon was predictive, and Gange was faithful to that original conception, so I don't think that criticism is fair.
but at first glance it appears he’s assuming the system in question is at equilibrium. If I’m correct, this makes his argument pretty uninteresting because life is inherently a nonequilibrium phenomenon, so it will obviously require nonequilibrium conditions to originate (and requires nonequilibrium conditions to continue). And of course, Earth is not and has never been at equilibrium….
I felt Gange had a lot of lose ends and probably erroneous inferences. I regarded his work better than most because it was the first creationist work I knew of that made a point of excluding classical thermodynamics from the question of design and instead appealing to the Generalized 2nd law. It was also the first creationist work that properly showed that entropy is not identified as disorder, that life is not ordered, etc. I'm not familiar enough with the Generalized 2nd law to say one way or another, but if that means in general uncertainty about the state of systems grows over time, or that somehow transmission of information is corrupted in the copying process, then his argument holds. Certainly in information processing, it is evident there is loss of fidelity in all transmissions. Even without appeal to thermodynamics it is apparent coordinated information (like we find in software) simply deteriorates as its transmitted, it doesn't get better or more coordinated and develop more complex capabilities. Formally demonstrating that is a good step for the ID community. Now, if you find holes in Gange's reasoning, I would be appreciative to hear because as I said, I have concerns that he was hasty with his conclusions. Even supposing his final conclusion is correct, his inferential method is open to criticism. Thanks for your insightful criticism.scordova
December 12, 2013
December
12
Dec
12
12
2013
11:37 PM
11
11
37
PM
PDT
Life will not spontaneously arise in a closed system. Gange asserted that the problem of a system spontaneously creating novel life-giving information is like the problem of perpetual machines.
Exactly. Random evolution is impossible as a perpetual motion machine. This is the thermodynamics argument contra evolution. Gange, Szilard, Sewell et al. express it with different words, but the substance is the same. Evolution is a Maxwell demon, i.e. something that cannot work at zero information/energy cost. It works gratis only in abstracto, in the mind of Maxwell/Darwin. If one tries to construct a real system working as Maxwell/Darwin demon one finds that it needs more information/energy than those produced, as thermodynamics states (and as Szilard computed). Maxwell’s demon links thermodynamics with information theory. As Dembski says: “It is CSI that enables Maxwell’s demon to outsmart a thermodynamic system tending toward thermal equilibrium” (Intelligent Design, pag. 159)
You're getting Szilard's point almost exactly backward; the entire point of his analysis is that an intelligent agent (the "imp") cannot decrease entropy. As far as the laws of thermodynamics are concerned, intelligent agents are subject to the exact same constraints as everything else. That's the point of Maxwell's Demon: it appeared to Maxwell that an intelligent agent could decrease entropy in violation of the second law; Szilard's analysis shows that the thermodynamic cost of thinking is high enough to compensate for the decrease, leaving the second law intact. Even if Szilard (and pretty much all subsequent analysis) was wrong, and Maxwell's Demon could decrease entropy in an isolated system, it wouldn't be by performing intelligent design; it would be by menially sorting molecules one-by-one. The information that's relevant isn't the sort that intelligent design is concerned with (CSI etc), but information that describes the precise state of a physical system: there's an atom there, this molecule is more energetic than that one, etc. This is not the sort of information that design produces; it is the kind that measurement produces. So, there's a link between thermodynamics and information theory, but its implications are completely different from what you're claiming. Furthermore, you're ignoring the fact that evolution happens (and presumably abiogenesis happened) in open systems with nonequilibrium boundary conditions. Dismissing this as unimportant simply shows ignorance of basic thermodynamics. Understanding how a system interacts with its surroundings is absolutely critical to understanding the thermodynamic constraints on it. Take the most obvious difference: in isolated systems, entropy always increases; in open systems, entropy decreases are not only possible but utterly commonplace. Compare: Isolated systems without intelligent agents: entropy always increases. Isolated systems with intelligent agents: entropy always increases (per Szilard's analysis). Open systems without intelligent agents: entropy often decreases. Open systems with intelligent agents: entropy often decreases. As far as thermodynamics is concerned, isolated vs. open matters, but intelligent vs. non-intelligent does not. BTW, a note for Sal: I haven't had a chance to go through Gange's argument properly yet, but at first glance it appears he's assuming the system in question is at equilibrium. If I'm correct, this makes his argument pretty uninteresting because life is inherently a nonequilibrium phenomenon, so it will obviously require nonequilibrium conditions to originate (and requires nonequilibrium conditions to continue). And of course, Earth is not and has never been at equilibrium.... Gange also misstates both the problem of Maxwell's Demon and the solution in significant ways: he says that the demon supposedly creates energy, but in fact (if it worked) it would decrease entropy (thus making the energy that was already there avalialble to perform work). He also says that the resolution of the demon paradox is that "it costs him energy to acquire the information he needs", but more recent analysis has shown that information can be gained "for free"; it's the cost of forgetting the leftover information ("there used to be an atom there") to make room for the next measurement that's costly. This doesn't bode well for Gange's understanding of thermodynamics. I'll try to get a chance to go through Gange's argument, but won't have time for at least a few days.Gordon Davisson
December 12, 2013
December
12
Dec
12
12
2013
10:57 PM
10
10
57
PM
PDT
Let me state my view in the previous comment more succinctly: Szilard's paper is an analogy (a metaphorical illustration) of the problems facing biological evolution, it isn't the problem facing evolution. No Free Lunch and various other ID arguments attempt to exorcise Darwin's demon like Szilard exorcized Maxwell's demon.scordova
December 12, 2013
December
12
Dec
12
12
2013
12:57 PM
12
12
57
PM
PDT
I should point out, I'm not averse to making analogies from thermodynamics to the problem of evolution. Where I have issues is equating (versus relating) thermodynamics and biological evolution. I tried to point out the complication in equating thermal information with organizational information in this essay: https://uncommondescent.com/computer-science/specified-entropy-a-suggested-convention-for-discussion-of-id-concepts/ If we look at Szilard paper as an analogy to the problem of information in Darwinian evolution, I'm all for it. Where I have issue is saying Szilard's paper directly disproves Darwinian evolution. The main reason is information referred to in Szilard's paper is only a specialized case of the more general class of information involved in the ID vs. Darwin debate. The paper was a landmark in physics, and it was a pre-cursor to the Law of Conservation of Information that is being argued by ID proponents today. I support the notion of Conservation of Algorithmic Information, but I've said CSI is not the same as Algorithmic Information. But that's a separate confrontation (ahem, I mean, discussion). :-) IMHO, Szilard's paper deals with an information problem that is far more difficult to understand than the information problem in Darwinian evolution. The information problem for Darwinism is much easier to describe as I tried to illustrate here: https://uncommondescent.com/computer-science/dawkins-weasel-vs-blind-search-simplified-illustration-of-no-free-lunch-theorems/ Appeals to Maxwell's demon might actually add an unneeded layer of difficulty in understanding the ID debate. But it should still be discussed because it is a related topic and it serves the ID community well to weigh all information related to ID. If Szilard paper helps someone understand ID better, more power to them. For me, the paper was interesting in its own light, not to mention the utterly captivating title that had the phrase "Intelligent Beings". :-)scordova
December 12, 2013
December
12
Dec
12
12
2013
12:22 PM
12
12
22
PM
PDT
For those wanting to understand Maxwell's demon as well as its relationship to ID, hopefully the following essay will clarify as it links to a great video on the topic. The video is superb, and then I added a few remarks to relate the problem to ID. https://uncommondescent.com/physics/csi-and-maxwells-demon/ Hope this helps.scordova
December 12, 2013
December
12
Dec
12
12
2013
11:23 AM
11
11
23
AM
PDT
Life will not spontaneously arise in a closed system.
Nor in an open system (as Niwrad points out). There is no substantive distinction between a closed system and an open one. Just linguistic semantics.Eric Anderson
December 12, 2013
December
12
Dec
12
12
2013
08:27 AM
8
08
27
AM
PDT
RE: JGuy post 6 I'm not a physicist, but I think what Szilard has shown (or is otherwise purported to have shown) is that the imp's operation of the window was not "effortless" (or as the original OP lists "The imp is able to open or close the window at will, and without effort. "). My understanding of the OP is that Szilard said that the imp would need to first identify which air molecules are fast moving and which are not. What constitutes as a "strong breeze" or a "strong enough breeze" (rephrased) to indicate to the imp to raise the window. I'm still stuck on how high pressure air could develop in one of the two boxes in the first place! Seems like the box with the imp would have the highest pressure would it not? His physical presence coupled with the additional physical presence of the shelf would cause a reduction in the physical volume of the box available for the gas to occupy. A "squeeze" if you will. The gas (air) would essentially need to to pressurize however more slightly than the other box by virtue of the physical constraints of the box. I think Szilard was suggesting that the imp would operate as a pressure release valve essentially. That is once there is enough energy (pressure) built up (and I'm not sure how the build up would occur...maybe the imp's biological processes curse him with flatulance?), the imp would open a window. However, just as with a pressure release on a boiler, some energy is lost because it performs work to push open the valve (sometimes a coil spring or other stiff yet flexible metallic piece). I think in the case of the imp, the energy is consumed by the imp or otherwise converted into a form of information such that the imp recognizes the "strong breeze" and opens the window. Am I misunderstanding Szilard's solution?ciphertext
December 12, 2013
December
12
Dec
12
12
2013
07:39 AM
7
07
39
AM
PDT
Life will not spontaneously arise in a closed system. Gange asserted that the problem of a system spontaneously creating novel life-giving information is like the problem of perpetual machines.
Exactly. Random evolution is impossible as a perpetual motion machine. This is the thermodynamics argument contra evolution. Gange, Szilard, Sewell et al. express it with different words, but the substance is the same. Evolution is a Maxwell demon, i.e. something that cannot work at zero information/energy cost. It works gratis only in abstracto, in the mind of Maxwell/Darwin. If one tries to construct a real system working as Maxwell/Darwin demon one finds that it needs more information/energy than those produced, as thermodynamics states (and as Szilard computed). Maxwell’s demon links thermodynamics with information theory. As Dembski says: "It is CSI that enables Maxwell’s demon to outsmart a thermodynamic system tending toward thermal equilibrium" (Intelligent Design, pag. 159) I would add that life spontaneously arise not even in a open system, if intelligence doesn't intervene.niwrad
December 12, 2013
December
12
Dec
12
12
2013
03:49 AM
3
03
49
AM
PDT
JG: The idea is, that by allowing faster molecules to pass one way and slower ones the other, Maxwell's Demon was creating a pressure difference -- and a temperature difference. Temp is related to avg speed of molecules, and pressure is related to avg momentum transferred in collisions with the walls of a container. From these work could be extracted. However, to know which was fast or slow, some observations have to be made and that requires work. It turns out, you will not get something for nothing. (This is yet another case of intelligently directed ordering work that undoes the overwhelmingly expected result of diffusion.) KFkairosfocus
December 11, 2013
December
12
Dec
11
11
2013
11:10 PM
11
11
10
PM
PDT
The imp didn't actually move anything. The window is all, and that was supposed to be done effortlessly. I suppose what I'm suggesting is that this final energy that turns some turbine was already there. It is the energy in the gas (fast flying gas particles). By segregating the fast from slow gas particles (atoms or molecules) into two compartments merely by allowing the random pressure difference to enter one compartment doesn't seem, to me yet, to create energy. The particles already have the kinetic energy, it seems to me the only thing occurring is that it is being isolated to be made useful at a macroscopic scale. Or in other words... All around us are molecules in motion. This kinetic energy is measured and called some temperature. Therefore, the energy is already there. The particles are so numerous and so small that we can not practically utilize this kind of energy. All the imp did, so far as I see, is effortlessly separated the faster particles from the slow to allow for that existing energy to be useful at a macroscopic scale. Let me simplify this with an extreme example. Suppose you had two chambers as described above. With a small window... let's suppose there are only two gas molecules flying around in these chambers. Let's suppose one gas molecule is fast and the other slow - not that this really matters. It seems very likely that you will have varying configurations of gas molecule locations relative to the compartments. Sometimes both molecules are in one container, sometimes separate...all due to chance! ... So, when the fast molecule moves into the same chamber as the slow one, was energy created? or not? After-all, when they are both in one container, the other is a vacuum. I think you can see why I would argue no energy was created...nor was it an enigma that one compartment was hotter than the other.JGuy
December 11, 2013
December
12
Dec
11
11
2013
09:58 PM
9
09
58
PM
PDT
It doesn’t seem to me that energy is really created by the imp. It seems moreso that available energy is harnessed and made useful. The generator spinning in the window will slow down molecules flowing back into the slow compartment. That lost energy is captured into the form of electrical energy. So, where is the net increase? What is wrong with my thought?
Pressurized gas spontaneously moves into a vacuum. This movement can be used to drive a generator. The imp could move gas from a container and make it a vacuum. Szilard was able to show the imp needed energy to locate the atoms.scordova
December 11, 2013
December
12
Dec
11
11
2013
04:29 PM
4
04
29
PM
PDT
I just updated the OP to include a photo of Szilard with Einstein and provided some biographical information on him. Wow! The guy was first rate!scordova
December 11, 2013
December
12
Dec
11
11
2013
04:25 PM
4
04
25
PM
PDT
Nice Post. Lots of stuff to digest.Mapou
December 11, 2013
December
12
Dec
11
11
2013
03:03 PM
3
03
03
PM
PDT
p.s. What if the window is left open with no demon and a breeze blows a little more into one room than the other? At that moment, was energy created (without a need for an imp)?JGuy
December 11, 2013
December
12
Dec
11
11
2013
02:03 PM
2
02
03
PM
PDT
It doesn't seem to me that energy is really created by the imp. It seems moreso that available energy is harnessed and made useful. The generator spinning in the window will slow down molecules flowing back into the slow compartment. That lost energy is captured into the form of electrical energy. So, where is the net increase? What is wrong with my thought?JGuy
December 11, 2013
December
12
Dec
11
11
2013
02:01 PM
2
02
01
PM
PDT

Leave a Reply