News for hackers. Stuff that matters.
Re:Maxwell’s Demon Continues to Startle Scientists by scrooched_moose on Thursday May 20, 2021 @10:16PM (#27229806)
Possessing only an undergrad level of physics, I've never understood why this is a remotely interesting question.

It seems to effectively be equivalent to "Imagine we have a refrigerator with a massless, frictionless motor that also doesn't need to be plugged in. Isn't it crazy that this refrigerator would violate the second law of thermodynamics?"

Re:Maxwell’s Demon Continues to Startle Scientists by ineptech on Thursday May 20, 2021 @10:39PM (#27229933)
A motor that doesn't require energy violates the second law by definition; a selectively-enforced barrier that lets hot particles through but not cold particles didn't obviously provably cost energy.

If that sounds like a cop out, remember that the Demon is a metaphor that's supposed to represent an as-yet-undiscovered technology. Like, instead of a literal massless door that opens and closes, something like, "What if we discovered a clever way of designing a little gap between two chambers and ionizing the particles such that only hot ones go through."

Re:Maxwell’s Demon Continues to Startle Scientists by thaumasiotes on Friday May 21, 2021 @02:58AM (#27231250)
> a selectively-enforced barrier that lets hot particles through but not cold particles didn't obviously provably cost energy.

I think you're leaving something out of your description. We have passive selective barriers that will admit low-energy particles while reflecting high-energy particles. (Cotton, say, will reflect visible light while being transparent to infrared light.)

So what do you get by hanging a sheet between two rooms? Nothing.

I think Maxwell's Demon will only let hot particles pass through in one direction, but that's nothing new either; we have one-way mirrors.

Re:Maxwell’s Demon Continues to Startle Scientists by Dylan16807 on Friday May 21, 2021 @03:26AM (#27231387)
> I think Maxwell's Demon will only let hot particles pass through in one direction, but that's nothing new either; we have one-way mirrors.

We don't, actually. We have semi-transparent mirrors that allow a small fraction of the light through, and they appear "one-way" only when one side is vastly brighter than the other.

Re:Maxwell’s Demon Continues to Startle Scientists by jayd16 on Friday May 21, 2021 @03:34AM (#27231426)
If one way mirrors worked like that we wouldn't need heat pumps. We do because they don't work like that.
Re:Maxwell’s Demon Continues to Startle Scientists by hedora on Thursday May 20, 2021 @10:59PM (#27230019)
Imagine the door had some mass and some friction, but the particles were really heavy, so even a few correct openings could create a large energy imbalance.

There’s no apparent relationship to the energy gain per separated particle and the energy used by the door. That makes it hard to prove it’s not generating energy.

That’s what makes it an interesting problem.

Re:Maxwell’s Demon Continues to Startle Scientists by scrooched_moose on Thursday May 20, 2021 @11:13PM (#27230083)
Interesting, thanks. I hadn't considered the disconnect between the particles and the door.
Re:Maxwell’s Demon Continues to Startle Scientists by lmilcin on Friday May 21, 2021 @05:32AM (#27232023)
This approach is assuming thermodynamics as a unquestioningly true law.

That could be fine for a student that is learning physics but not for a scientist.

Treating something as unquestioningly true is a sign of belief and religion.

Now, for most everyday situation it is practical to assume that a device can't be creating energy from nothing, can't be ordering atoms in two containers without using energy, because we know there exists second law of thermodynamics.

But Maxwell's Demon is an experiment specifically aimed at testing thermodynamics.

We can use the second law of thermodynamics exactly because we expend/expended effort to test it to a sufficiently high degree and we are revisiting as we learn new facts and push the boundary.

Re:Maxwell’s Demon Continues to Startle Scientists by libeclipse on Thursday May 20, 2021 @11:59PM (#27230351)
I agree but for a different reason. It seems obvious that Maxwell's demon would itself use energy, thereby paying for the "decreased entropy" with this energy.

This is essentially the same as:

> The second vital piece of the puzzle was the principle of erasure. In 1961, the German American physicist Rolf Landauer showed that any logically irreversible computation, such as the erasing of information from a memory, would result in a minimal nonzero amount of work converted into heat dumped into the environment, and a corresponding rise in entropy.

Re:Maxwell’s Demon Continues to Startle Scientists by exporectomy on Friday May 21, 2021 @05:00AM (#27231856)
Why is it obvious that Maxwell's demon would use energy? It could be a frictionless spring-loaded door with latches to hold it open or closed and it bounces between states when it's unlatched.
Re:Maxwell’s Demon Continues to Startle Scientists by tomcam on Friday May 21, 2021 @01:46AM (#27230900)
I wish someone would address your point. This leapt to my mind immediately but I don’t understand enough about the physics to go any further with it.
Re:Maxwell’s Demon Continues to Startle Scientists by TchoBeer on Friday May 21, 2021 @08:37AM (#27233234)
iirc Laundauer assumes the second law, so it's not a good refutation here.
Re:Maxwell’s Demon Continues to Startle Scientists by canjobear on Thursday May 20, 2021 @10:31PM (#27229890)
The interesting thing is that this theoretically perfect refrigerator operates in a way that hints at deep connections between intelligence, information, and entropy.
Re:Maxwell’s Demon Continues to Startle Scientists by selestify on Friday May 21, 2021 @02:00AM (#27230969)
What’s the connection between intelligence and entropy?
Re:Maxwell’s Demon Continues to Startle Scientists by a1369209993 on Friday May 21, 2021 @02:50AM (#27231210)
"Intelligence" is a little misleading, it's more along the lines of "awareness of one's environment".

Per Liouville's Theorem, information about your surroundings is a conserved quantity. A physical system (persistent along a time-like curve) that has more information about a particular part of it's environment (eg "the content of this movie is so-and-so") in one light cone of a event than the other, must have correspondingly more information about some other part of it's environment (eg "this heat sink that I am erasing scratch space with has particle velocities within bounds such-and-such") in the other light cone compared to the first one.

If one light cone has much larger quantities of compressible ('cold', kind of) information than the other, such that interesting information about our surrondings always appears on the opposite side of events from it, we call that light cone "the past".

I don't know that we have a better explanation than the anthropic principle for why one one of our light cones is like that, though.

Re:Maxwell’s Demon Continues to Startle Scientists by thaumasiotes on Friday May 21, 2021 @03:02AM (#27231279)
> we call that light cone "the past"

Hmmm. I had the impression that the flow of time was implicit in the fact that we talk about "light cones" instead of "light spheres". Is that not the case?

Re:Maxwell’s Demon Continues to Startle Scientists by a1369209993 on Friday May 21, 2021 @03:16AM (#27231338)
Given well-behaved (ie relatively flat-ish) spacetime, there's a distinction between light cones going this way, and light cones going that way (a particle is a graph[0] edge that is attached to this-ward side of one event and the that-ward side of another), but general relativity, quantum mechanics, and particle physics - assuming you also handle the charge and parity from CPT symmetry - don't distinguish which of 'this' versus 'that' is past or future.

0: Quantum field theory makes this more complicated, since particles are more like amplitude distributions over possible graph edges, except events also aren't vertices but distributions over vertices, and if you can figure out a way to explain this that actually makes sense you might win a nobel prize, but it empirically is well aproximated by a graph at macro scales.

Re:Maxwell’s Demon Continues to Startle Scientists by prof-dr-ir on Friday May 21, 2021 @06:15AM (#27232246)
I would imagine that a theoretical physicist on LSD would sound like these comments. I love it.

But more seriously (and with tons of respect, because these topics are difficult) - why invoke all these complicated concepts? When we discuss statistical mechanics it is kind of understood that the ferromagnet we try to describe is not placed on the innermost stable orbit around a black hole, and similarly when I look in a mirror there is little need to worry about the difference between P and CPT.

I guess one can say that the essence of a physical system is not always in the fundamentals of physics?

Having said that, I do not know what "intelligence" is, so I do not know how to relate it to entropy.

Re:Maxwell’s Demon Continues to Startle Scientists by a1369209993 on Friday May 21, 2021 @01:35PM (#27237176)
> I would imagine that a theoretical physicist on LSD would sound like these comments. I love it.

I don't need LSD!

> When we discuss statistical mechanics it is kind of understood that the ferromagnet we try to describe is not placed on the innermost stable orbit around a black hole

If your theory of statistical mechanics requires that, then your theory is wrong, or at least badly incomplete. That might not be a problem you need to care about for practical reasons, but it is a problem.

> I guess one can say that the essence of a physical system is not always in the fundamentals of physics?

Depends on what you mean by "essence" and "fundamentals" - you're not going to have good time trying to analyze a computer [0] running a game-of-life sim in terms of it's constituent quarks, obviously - but there's usually something useful to be learned from the underlying physical laws.

0: (whose primary descriptive quality is that it's)

Re:Maxwell’s Demon Continues to Startle Scientists by ffhhj on Friday May 21, 2021 @02:05AM (#27230986)
information
Re:Maxwell’s Demon Continues to Startle Scientists by exporectomy on Friday May 21, 2021 @03:38AM (#27231450)
That refrigerator would still use energy compressing the refrigerant. That's Work=F*x energy which won't go to zero with a massless frictionless motor.

Or do you mean you also have a generator to recover the energy from the temperature differences it creates? So it's not really doing refrigerating because you're using all the cooling power to run the motor?

Re:Maxwell’s Demon Continues to Startle Scientists by Sosh101 on Friday May 21, 2021 @05:17AM (#27231934)
It's interesting because it's one of the intersections between physics and information theory. Why it's not possible is not at all obvious.
Re:Maxwell’s Demon Continues to Startle Scientists by bb88 on Thursday May 20, 2021 @10:22PM (#27229841)
Entropy can't be reversed, even if you have a massless, frictionless, motor that doesn't need to be plugged in. That is effectively what Maxwell's demon is.
Re:Maxwell’s Demon Continues to Startle Scientists by tsimionescu on Friday May 21, 2021 @01:43AM (#27230882)
Of course entropy can be reversed, I literally do it in my sleep!

However, reversing entropy requires an amount of energy commensurate with the amount of entropy you are reducing. Maxwell's demon would appear as a violation of this.

A massless, frictionless motor could trivially reverse all the entropy in the universe given enough time, but it is not clear that Maxwell's demon REQUIRES something equivalent.

Re:Maxwell’s Demon Continues to Startle Scientists by bb88 on Friday May 21, 2021 @12:53PM (#27236569)
Maxwell's demon requires erasure which would subsequently increase entropy.

https://www.nature.com/news/the-unavoidable-cost-of-computat....

Re:Maxwell’s Demon Continues to Startle Scientists by tsimionescu on Friday May 21, 2021 @01:31PM (#27237119)
As someone else was noting, this result is based on Landauer's limit, which is itself ultimately based on the second law of thermodynamics. If Maxwell's demon existed, the second law would be wrong, so Landauer's limit would be wrong, so there could be a way to erase information without increasing entropy.

I'm also not sure why Maxwell's demon would require erasure.

Re:Maxwell’s Demon Continues to Startle Scientists by bb88 on Friday May 21, 2021 @05:00PM (#27239909)
Wikipedia summarizes this well.

https://en.m.wikipedia.org/wiki/Maxwell%27s_demon

I think it's clear that by now the 2nd law always wins. Always.

Re:Maxwell’s Demon Continues to Startle Scientists by dw-im-here on Friday May 21, 2021 @02:22PM (#27237887)
Because it's a way to investigate the properties of information and entropy