From: sbharris@ix.netcom.com(Steven B. Harris) Newsgroups: sci.physics Subject: Re: Teleportation? Date: 26 Nov 1998 19:14:39 GMT In <365D754F.175E@ix.netcom.com> Uncle Al <UncleAl0@ix.netcom.com> writes: > >Gil Hagi wrote: >> >> Hi, I'm doing my chem ISP on the feasability of teleportation. I was >> just wondering what thoughts you people might have about how possible >> this is, what potential hurdles might exist in making this possible. >> From what I see, the only physical problems with teleportation are >> heisenberg's uncertainty principal (which I hear someone has found a >> way around), and computing power (which will be there in a few >> years...) >> >> So what else is there? >> >> Gil. > >If you are going to transport the substance, figure out he E=mc^2 energy >of 70 kg of human being. If you are going to transport the information... >how many bytes are we talking? Read some Claude Shannon and note a mole >contains 6x10^23 little fellas. The nice thing is that a lot of info is redundant. Your body contains a lot of water molecules (in fact numerically MOST of the body's molecules) which can go anywhere-- only vague info on placement needed, and none at all on orientation. Same for a lot of generic chemicals in solution. Proteins? You only have 100,000 different ones or so (same number as the number of your genes) and once you identify one, it takes only ln(100,000) = 6 x ln10 = 14 bits to say what you've got (assuming modest code library on receiving end). Add a few more for spacial orientation (say 6, giving 32 possible solid angles for the agreed "head" to point in) if it's a lipid-bound protein, and 2 x 9 x ln10 = 40 bits or so for the nanoassemblers to place it on the 1 meter square 2-D receiver plate (again, assuming a code library grid for receiver placement). Think of it as DVD for humans. How many protein molecules in a human? Figure 100,000 dalton molecules (1.7e-21 kg) and 10 kg of protein and you get 6e21 molecules, each specified (if you give each a number like a stone in some old cathedral being moved from here to there) by 50 bits. Add it up for proteins and you've got 6e21 molecules to place, each requiring 14 + 6 + 40 = 60 bits of info. Total of 4e22 bits, perhaps, and that's a pretty good start. Present fiberoptic info transfer rates are only 1e11 bits/sec or 3e18 bits/yr. So it would 10,000 yrs to transfer a human. But Moore's law works for this kind of thing, also, and you get a factor of 10 improvement every 8 years, with no obvious law of physics in the way for a long while. That means in 40 years we should be able to transfer this in 5 weeks or so. People will be faxed and photocopied, but not like on Star Trek. Instead, the human in metabolic statis will be grown out of a 2-D replicator plate (on which layers are assembled by nanomachines) over the space of weeks to months. Not the greatest for business convention travel, but a good way to get to Pluto or that earthlike planet around Alpha Centauri A, maybe. Also good for cryonics, or backup copies in case you get cremated by a volcano, or something. Whether this is "you" or not is a semantic, not a scientific question. From: sbharris@ix.netcom.com (Steven B. Harris) Newsgroups: sci.physics,sci.physics.research Subject: Re: Downloading hardware from the Internet Date: 26 May 1999 00:16:40 GMT In <7i3nap$kov@news.tue.nl> Karel Knechten <karelk@stack.nl> writes: >I think it will be impossible to transport everything [as >information], unless one can build it up from it's atoms - and that's >impossible because of Heisenberg. > >Greetinkz, >Karelk Heisenberg keeps you mostly from seeing where electrons are-- it doesn't keep you from locating atoms reasonably well-- we'll all seen the AFM and STEM pics. If you have a bunch of little machines that feel out surfaces and identify atoms and send the positional information to a bunch of other machines which assemble atoms from a stock into an idential 2-D surface, you're off an running. You can tear down a 3-D object on one side of your transmitter, and build it up on the other side, an atomic or molecular layer at a time. You might have to keep the object quite cold while the process goes on, but for many objects that's not a problem. How fast? Drexler of the Nanotechnology crowd has done some prelimary calculations on how fast machines with arms made of tubes of carbon atoms might be able to move them, and it's quite fast-- millions of times a second. How fast you can put down a layer will depend on how much information is in it, just as tiling a floor with some jigsaw puzzle scene takes longer than a simple pattern. So crystals will go fast, and biological specimens will go slowly. It won't be like Star Trek, but more like watching bamboo grow, I suspect. Biological specimens might not go as slowly as you think, though, since there isn't as much information even in an organism as first appears. If a human has 100,000 genes for that many proteins, you need only send the structure for each of 21 amino acids once, then the full amino acid code for each of the 100,000 proteins once, and after that the machines only have to look at the protein well enough to distinguish it from all others (which can be done by reading only a few amino acids), and then the entire molecule can be ripped out at the transmitter end, and the data sent as one 16-bit code number with a tag denoting a protein found, to the other. At which point the assembler gets the appropriate protein molecule from the store, and tacks it down to the spot. And so on. A few more bits may be needed for orientation of some proteins. Water molecules are generic and a lot of lipids also. Orientation will matter only for things found naturally in organisms tacked to a solid phase, and even then, only such orientation which is expected to be resistant to thermal buffeting in the live organism, which isn't much. Since cells and small clusters of cells (embryos) can be recovered from liquid helium now, there is not problem in sending such things by matter trasporter, and having them arrive "alive" (revivable). For larger collections like organs and organisms the technology awaits only good methods of cryopreservation which obviate chilling injury and ice damage (which happens now mostly between cells and due to osmotic forces, contrary to a lot of SF hoopla). It should be possible to replace enough water in organisms with organic solvents that little ice is formed on solidification, and most of the liquids vitrify. This may allow cryopreservation of organs and even people with immediate viability (true suspended animation). Matter transmission of vitrified organisms, including people, ought to be possible. FAXing and Photocopying of people is not a physical impossibility, so far as I can tell-- it's just a matter of technology. For some of the philosophical implications, one can see Algis Budrys' _Rogue Moon_ (1960) and many other SF tales that have followed. What makes you YOU, and what makes a repaired vitrified you, YOU, and what makes a copy, or a transmitted version of you, YOU? These are more value judgements than meaningful physics questions. They don't have objective answers. The question of identity is the problem of the Sorites paradox, also known as the Thesius's ship problem (replace a ship a board at a time, and when does it cease to become the original ship?). We are information beings, and it's the information that makes us who we are, not the atoms. But how much information is critical to identity? How is it that quantitative changes finally add up to a qualitative change? We have no answer for that because qualitative changes are usually in the mind's eye. But it'll be fun to grapple with when the time comes. Steve Harris |