∫ Of Physics and Other Demons

This is the most interesting physics paper I’ve seen in a while. Technically it’s not out yet; it’s in preprint now, which is why I first found it on the ArXiv blog, and why you probably can’t follow the link I left you unless you’re a scientist or something. Sorry about that.

Anyway, This group in Japan (principally at Tokyo University) has built a real-world version of something originally proposed by Leó Szilárd* back in 1929. Szilárd’s idea was a thought experiment that showed how information could, theoretically, be used to store energy.  His basic idea was this: Imagine a box with a valve of some kind (for simplicity, imagine it’s a force-field that you can raise and lower at will. Then, put an electron in the box, and heat it up so the electron bounces around a bit. Then raise the force-field; the particle will be trapped on one side or the other, like so:

Now, if you know which side of the force field the electron is on, you can attach a piston to the other side, and push it in. Meanwhile, the electron is still bouncing around randomly, because of it’s thermal energy.

Note that it didn’t take you any energy to push the piston in, because there’s nothing in the other side. In fact, if the piston was there all along, the vacuum inside the chamber will pull it in on its own. But at that point, imagine what happens when you lower the force field: All of a sudden, the electron is free to start bouncing off the piston, which will gradually push it back to the left. But when a piston moves to the left, that’s called “work.” It’s useful energy you’ve produced. You could put a little car next to the piston and it will get a little push! And you seem to have gotten this energy for free.

Szilárd’s great idea came from explaining why  you can’t actually get this energy for free. Notice that, in order for this to work, you have to know which side of the box the electron is on. You could take a random guess, but if you do this enough times, you’ll guess wrong about half of the time, in which case you’ll be trying to push the piston into a region where the pressure suddenly got twice as large. So it’ll be pushed out instead, and you’ll lose the same amount of energy you might have gained**.

But, if you could always get right, you can keep producing energy. In fact, it’s as if, Szilárd discovered, you had made the energy directly from that bit of information you had about where the particle was.

Neat idea, but seems a bit far fetched. Or it did, until these guys in Japan did it for real. They used a slightly different arrangement, but they pulled off the same concept of converting information into energy. Note that they still haven’t made energy from nothing; it takes you some energy to get that information in the first place. Still, the fact that you can basically store this energy in the form of information and get it back later is pretty cool. It’s like an information Battery.

The Tokyo group’s apparatus consisted of a set of tiny “stairs” and a tiny little ball. Not as small as an electron, but still small enough that the random wiggling of the air around it makes it bounce around a bit. Sometimes, it will even bounce enough to make it up to the next stair step; when it does, you’ve got a bit of energy you can use there  (you may not see this, but notice: a brick lying at the foot of a stairway won’t do anything useful, but if I have a brick that’s up a few stairs, I can imagine doing something useful with that, like dropping it on a water wheel or something).

Anyway, in general this doesn’t help us, because the random motion of the little ball will just as often make is slip back down a stair as it will make it climb. However, if we had the right information, we could throw up a force-field behind the ball every time it climbed a step, so that it couldn’t slip back down.

If you do this over and over again, you can get it all the way up to the top of the stairs. And then do something useful with it, getting energy out of it by dropping it off the top. A little bit like these penguins:

Only the penguins are doing this using the energy stored in a real battery; and the Tokyo group does it with energy just stored in the form of information. They use a little video camera to monitor the position of the bead, and every time it moves up a step, they know to move the force field up a step, too. A pretty picture I got from the ArXiv blog demonstrates this nicely:

By  the way, these folks have drawn a little “demon” moving the force-field around as a reference to a famous thought experiment called “Maxwell’s Demon,” which served as the groundwork for Szilárd’s idea. But that’s a bit outside the scope of this discussion.

Anyway, before I leave you, let me make sure I’ve been clear about what’s going on here, and why it’s cool. These guys still haven’t gotten energy from nowhere; that’s impossible. They have to use energy to run the video camera. But that camera gets them information, which they can turn back into energy by making their little ball bounce up the stairs. The energy they put into the camera was stored in the information before it was released back into the bead system. This may seem weird, but this has to be the case. Otherwise, where else could the energy come from that’s moving the ball up the stairs?


Right now, of course, there’s no practical use to this. Who wants to try to run their car off of a tiny little staircase? But just think what this means in concept: if information can be used as a place to store energy, it can be used as a way to move it around. And information is much, much easier to move around than just about any of the ways we have to store information right now.

Here’s a simple thought-experiment extension of this idea: We didn’t gain any energy from the staircase because we had to use the energy to run the camera. But all we needed the camera for was information. What it we put the camera up in a satellite, looking down at the staircase through a telescope (bear with me here. I know it’s too small to see, but it could work in principle). Then, we could easily power the camera with solar energy– the kind that’s very plentiful our in outer space, but harder to harvest here on Earth. We can also power a little radio transmitter on the satellite with this solar energy, and the camera can transmit the signal to move the forcefield every time it sees the ball move. We can get the same effect, only as far as people watching the stairs down on Earth are concerned, the energy is coming from nowhere. Because it’s being stored up as information and radioed down to us from outer space.


There’s actually a complex mathematics that’s evolving to describe how information can be used this way. In cutting-edge theoretical physics, people are beginning to discover that energy, entropy, and information are all bound up more tightly than we would have thought. Perhaps they’re even all sides of the same coin. But there are some fascinating subtlties. For example, notice that, just like a battery, the information doesn’t “go away” once you have used it to move the ball. It just ceases to be useful. Knowing the ball has moved from step 1 to step 2 will help you get energy once, but once it’s on step two that knowledge is just like a dead battery. Only in a dead battery, we can explain scientifically what has changed from the time the battery was still active. Should there be a way to tell if information has any useful energy stored in it?

I don’t have the answer to that. But it makes me want to study information theory.


*Szilárd is more famous for a different sort of contribution he made to the world a few years later: you probably know the story about how Albert Einstein wrote a letter to FDR which convinced him that an atomic bomb was a real possibility, leading directly to the Manhattan project. What you probably don’t know, however, is that Einstein didn’t personally write that letter. It was drafted by Szilárd, at the time a somewhat unknown scientist, who asked Einstein to send it to the president under his name so that it would be taken more seriously.


**I’m cheating a bit on the explanation here, because I’m making it sound like sometimes, by luck, you could get energy for nothing, and that you only break even in the long term. Obviously that isn’t right. Unfortunately, to really account for all the energy in the problem things get a little complicated, but you can probably puzzle it out if you’re looking for a challenge. Keep track of what happens to the pressure in the various sides of the box at various times, and remember that even having a single electron in there constitutes a pressure.


About Colin West
Colin West is a graduate student in quantum information theory, working at the Yang Institute for Theoretical Physics at Stony Brook University. Originally from Colorado (where he attended college), his interests outside of physics include politics, paper-folding, puzzles, playing-cards, and apparently, plosives.

2 Responses to ∫ Of Physics and Other Demons

  1. Paul West says:

    Am I missing something, or doesn’t work also need to be done to maintain the force-field and move it each time the ball moves in the desired direction?

    • Colin West says:

      No, I think you’re right about that, and I was worried about it too when I first saw the paper. But I think the point is that, while the energy invested in the force fields may make this “information engine” a really inefficient engine, it doesn’t change underlying point that these scientists seem to have been able to transfer energy stored as information. After all If you just stop looking at the camera and move the force fields randomly, you use the same amount of energy, but don’t get anything out of it. The force-field energy is an incidental expense, but doesn’t go directly into the system.

      As a sort of rough analogy, I think it’s like the fact that you have to put a battery in you car in order to run it, even though the car doesn’t run on battery power. You know the majority of the energy running the car comes from the gasoline, because if you take the gas out of the system, you don’t get any energy just from the spark. In the same way, moving the force fields may be a required energy investment, but it isn’t where the extra energy is coming from, because if you do it without processing any of the information then nothing special will happen.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: