3 ms·

I recall having an idea at least a bit similar to this years ago when I was heavily studying evolutionary informatics. Some particularly avant garde types in t

by api 10y ago

I recall having an idea at least a bit similar to this years ago when I was heavily studying evolutionary informatics. Some particularly avant garde types in that field have posited that the universe rather than having two constituents -- matter and energy -- has three primary first-order constituents. The third is information. Information is not merely an epiphenomena of matter and energy but a primary "thing." If that is the case then there should be an E=mc^2 type equation that relates matter to information and energy to information and all three should be interconvertible. It would then further follow that energy can be converted into information and vice versa in the same way that matter can. I then imagined a Dyson swarm of solar power satellites that produce a data stream encoding the energy they collect. This stream can be subscribed to and decoded to reconstitute this energy remotely. To globally conserve energy there would have to be a two-way aspect to this -- I imagined the receiver of energy transmitting "challenges" to the swarm that are then "solved" to yield energy stored in the form of the solution. The receiver then receives these solutions and executes them to generate what in effect would look like local perpetual motion. (But in reality energy is still being conserved.) It would look like a cryptographic hashcash-style challenge-response system with proof of work, but the energy input of the POW function can be reversibly extracted elsewhere. If such a thing were possible and sufficiently efficient and could function in the presence of high latency, this could power a starship among many other things. If it were latency-tolerant it might also be a way to store energy. Save your laptop's power to its hard drive. It reminds me a little bit of the "telematter stream" propulsion system from Peter Watts' Blindsight.

eli_gottlieb 10y ago
>Some particularly avant garde types in that field have posited that the universe rather than having two constituents -- matter and energy -- has three primary first-order constituents. The third is information. Information is not merely an epiphenomena of matter and energy but a primary "thing." Is that really so avant-garde? It seems a bit weird to claim that thermodynamics and the arrow of time are epiphenomena when we can measure the relevant quantities experimentally. The more avant-garde thing seems to be the kinds of papers where they claim space-time or gravity are somehow emergent from entropy/information. >If that is the case then there should be an E=mc^2 type equation that relates matter to information and energy to information and all three should be interconvertible. Uhh, why? Even the more exotic physics-of-information work seems to end up equating entropy with space-time or something like that: mass-energy would bend space-time-entropy, but wouldn't be convertible into it. Admittedly, this is all way above my head.
logfromblammo 10y ago
If you could figure out how to remotely power a space habitat with nothing but downloaded prawn streams and lolcats, you, good sir, will end up ruling the galaxy, whether you want to or not. If useful quantum information-to-energy conversion requires an observer, you will also solve the planet's unemployment problem.
paulmd 10y ago
> Information is not merely an epiphenomena of matter and energy but a primary "thing." If that is the case then there should be an E=mc^2 type equation that relates matter to information and energy to information and all three should be interconvertible. I'm not sure I agree with this premise, but one thing I've been musing about is that there's probably an information-theoretic lower bound for the amount of energy it takes to transmit a given quantity of data a given distance. You can pick different points along a curve (higher bandwidth/higher frequency signals need higher power to produce a given SNR, and lower-bandwidth/lower-frequency signals can produce the same SNR at a lower power) but there is some asymptotic energy limit there that you cannot beat unless you have an infinitely sensitive receiver. Anyway, what I'm going for here is that your unifying theorem there would probably be Shannon-Hartley, since that deals with the transfer (derivative) of information. The "noise" is whatever natural equilibrium opposes the transformation process, and your SNR dB is the equivalent of the rate constant in chemistry. It's all fairly useless without some idea of how we convert hashes back into energy of course. Without that, there is insufficient data for a meaningful answer.
- btilly 10y ago
  There is a minimum amount of energy required to flip a bit. See https://en.wikipedia.org/wiki/Landauer%27s_principle https://en.wikipedia.org/wiki/Landauer%27s_principle for details. This corresponds to a minimum amount of energy needed to transmit a given number of bits. We are nowhere near this limit. However it is an upper bound that guarantees that Moore's Law can't possibly continue for classical computing to the end of this century. (Part of the interest in quantum computing is that it has no theoretical upper limits at all. However this comes with some very weird restrictions.)

by api 10y ago

eli_gottlieb 10y ago

>Some particularly avant garde types in that field have posited that the universe rather than having two constituents -- matter and energy -- has three primary first-order constituents. The third is information. Information is not merely an epiphenomena of matter and energy but a primary "thing." Is that really so avant-garde? It seems a bit weird to claim that thermodynamics and the arrow of time are epiphenomena when we can measure the relevant quantities experimentally. The more avant-garde thing seems to be the kinds of papers where they claim space-time or gravity are somehow emergent from entropy/information. >If that is the case then there should be an E=mc^2 type equation that relates matter to information and energy to information and all three should be interconvertible. Uhh, why? Even the more exotic physics-of-information work seems to end up equating entropy with space-time or something like that: mass-energy would bend space-time-entropy, but wouldn't be convertible into it. Admittedly, this is all way above my head.

logfromblammo 10y ago

If you could figure out how to remotely power a space habitat with nothing but downloaded prawn streams and lolcats, you, good sir, will end up ruling the galaxy, whether you want to or not. If useful quantum information-to-energy conversion requires an observer, you will also solve the planet's unemployment problem.

paulmd 10y ago

> Information is not merely an epiphenomena of matter and energy but a primary "thing." If that is the case then there should be an E=mc^2 type equation that relates matter to information and energy to information and all three should be interconvertible. I'm not sure I agree with this premise, but one thing I've been musing about is that there's probably an information-theoretic lower bound for the amount of energy it takes to transmit a given quantity of data a given distance. You can pick different points along a curve (higher bandwidth/higher frequency signals need higher power to produce a given SNR, and lower-bandwidth/lower-frequency signals can produce the same SNR at a lower power) but there is some asymptotic energy limit there that you cannot beat unless you have an infinitely sensitive receiver. Anyway, what I'm going for here is that your unifying theorem there would probably be Shannon-Hartley, since that deals with the transfer (derivative) of information. The "noise" is whatever natural equilibrium opposes the transformation process, and your SNR dB is the equivalent of the rate constant in chemistry. It's all fairly useless without some idea of how we convert hashes back into energy of course. Without that, there is insufficient data for a meaningful answer.

btilly 10y ago
There is a minimum amount of energy required to flip a bit. See https://en.wikipedia.org/wiki/Landauer%27s_principle https://en.wikipedia.org/wiki/Landauer%27s_principle for details. This corresponds to a minimum amount of energy needed to transmit a given number of bits. We are nowhere near this limit. However it is an upper bound that guarantees that Moore's Law can't possibly continue for classical computing to the end of this century. (Part of the interest in quantum computing is that it has no theoretical upper limits at all. However this comes with some very weird restrictions.)