Teleportation and the Student: Impossibility As A Lesson PlanPosted: March 7, 2015 Filed under: Education, Opinion | Tags: authenticity, blogging, collaboration, community, design, education, higher education, human body, in the student's head, learning, lossy compression, science fiction, Sean WIlliams, Star Trek, students, teaching, teaching approaches, teleportation, teleporters, thinking, tools Leave a comment
Teleporters, in one form or another, have been around in Science Fiction for a while now. Most people’s introduction was probably via one of the Star Treks (the transporter) which is amusing, as it was a cost-cutting mechanism to make it easy to get from one point in the script to another. Is teleportation actually possible at the human scale? Sadly, the answer is probably not although we can do some cool stuff at the very, very small scale. (You can read about the issues in teleportation here and here, an actual USAF study.) But just because something isn’t possible doesn’t mean that we can’t get some interesting use out of it. I’m going to talk through several ways that I could use teleportation to drive discussion and understanding in a computing course but a lot of this can be used in lots of places. I’ve taken a lot of shortcuts here and used some very high level analogies – but you get the idea.
- Data Transfer
The first thing to realise is that the number of atoms in the human body is huge (one octillion, 1E27, roughly, which is one million million million million million) but the amount of information stored in the human body is much, much larger than that again. If we wanted to get everything, we’re looking at transferring quattuordecillion bits (1E45), and that’s about a million million million times the number of atoms in the body. All of this, however, ignores the state of all the bacteria and associated hosted entities that live in the human body and the fact that the number of neural connections in the brain appears to be larger than we think. There are roughly 9 non-human cells associated with your body (bacteria et al) for every cell.
Put simply, the easiest way to get the information in a human body to move around is to leave it in a human body. But this has always been true of networks! In the early days, it was more efficient to mail a CD than to use the (at the time) slow download speeds of the Internet and home connections. (Actually, it still is easier to give someone a CD because you’ve just transferred 700MB in one second – that’s 5.6 Gb/s and is just faster than any network you are likely to have in your house now.)
Right now, the fastest network in the world clocks in at 255 Tbps and that’s 255,000,000,000,000 bits in a second. (Notice that’s over a fixed physical optical fibre, not through the air, we’ll get to that.) So to send that quattuordecillion bits, it would take (quickly dividing 1E45 by 255E12) oh…
- Information Redundancy and Compression
The good news is that we probably don’t have to send all of that information because, apart from anything else, it appears that a large amount of human DNA doesn’t seem to do very much and there’s lot of repeated information. Because we also know that humans have similar chromosomes and things lie that, we can probably compress a lot of this information and send a compressed version of the information.
The problem is that compression takes time and we have to compress things in the right way. Sadly, human DNA by itself doesn’t compress well as a string of “GATTACAGAGA”, for reasons I won’t go into but you can look here if you like. So we have to try and send a shortcut that means “Use this chromosome here” but then, we have to send a lot of things like “where is this thing and where should it be” so we’re still sending a lot.
There are also two types of compression: lossless (where we want to keep everything) and lossy (where we lose bits and we will lose more on each regeneration). You can work out if it’s worth doing by looking at the smallest number of bits to encode what you’re after. If you’ve ever seen a really bad Internet image with strange lines around the high contrast bits, you’re seeing lossy compression artefacts. You probably don’t want that in your genome. However, the amount of compression you do depends on the size of the thing you’re trying to compress so now you have to work out if the time to transmit everything is still worse than the time taken to compress things and then send the shorter version.
So let’s be generous and say that we can get, through amazing compression tricks, some sort of human pattern to build upon and the like, our transferred data requirement down to the number of atoms in the body – 1E27. That’s only going to take…
years. Um, again. Let’s assume that we want to be able to do this in at most 60 minutes to do the transfer. Using the fastest network in the world right now, we’re going to have get our data footprint down to 900,000,000,000,000,000 bits. Whew, that’s some serious compression and, even on computers that probably won’t be ready until 2018, it would have taken about 3 million million million years to do the compression. But let’s ignore that. Because now our real problems are starting…
- Signals Ain’t Simple and Networks Ain’t Wires.
In earlier days of the telephone, the movement of the diaphragm in the mouthpiece generated electricity that was sent down the wires, amplified along the way, and then finally used to make movement in the earpiece that you interpreted as sound. Changes in the electric values weren’t limited to strict values of on or off and, when the signal got interfered with, all sorts of weird things happen. Remember analog television and all those shadows, snow and fuzzy images? Digital encoding takes the measurements of the analog world and turns it into a set of 0s and 1s. You send 0s and 1s (binary) and this is turned back into something recognisable (or used appropriately) at the other end. So now we get amazingly clear television until too much of the signal is lost and then we get nothing. But, up until then, progress!
But we don’t send giant long streams across a long set of wires, we send information in small packets that contain some data, some information on where to send it and it goes through an array of active electronic devices that take your message from one place to another. The problem is that those packet headers add overhead, just like trying to mail a book with individual pages in addressed envelopes in the postal service would. It takes time to get something onto the network and it also adds more bits! Argh! More bits! But it can’t get any worse can it?
- Networks Aren’t Perfectly Reliable
If you’ve ever had variable performance on your home WiFi, you’ll understand that transmitting things over the air isn’t 100% reliable. There are two things that we have to thing about in terms of getting stuff through the network: flow control (where we stop our machine from talking to other things too quickly) and congestion control (where we try to manage the limited network resources so that everyone gets a share). We’ve already got all of these packets that should be able to be directed to the right location but, well, things can get mangled in transmission (especially over the air) and sometimes things have to be thrown away because the network is so congested that packets get dropped to try and keep overall network throughput up. (Interference and absorption is possible even if we don’t use wireless technology.)
Oh, no. It’s yet more data to send. And what’s worse is that a loss close to the destination will require you to send all of that information from your end again. Suddenly that Earth-Mars teleporter isn’t looking like such a great idea, is it, what with the 8-16 minute delay every time a cosmic ray interferes with your network transmission in space. And if you’re trying to send from a wireless terminal in a city? Forget it – the WiFi network is so saturated in many built-up areas that your error rates are going to be huge. For a web page, eh, it will take a while. For a Skype call, it will get choppy. For a human information sequence… not good enough.
Could this get any worse?
- The Square Dance of Ordering and Re-ordering
Well, yes. Sometimes things don’t just get lost but they show up at weird times and in weird orders. Now, for some things, like a web page, this doesn’t matter because your computer can wait until it gets all of the information and then show you the page. But, for telephone calls, it does matter because losing a second of call from a minute ago won’t make any sense if it shows up now and you’re trying to keep it real time.
For teleporters there’s a weird problem in that you have to start asking questions like “how much of a human is contained in that packet”? Do you actually want to have the possibility of duplicate messages in the network or have you accidentally created extra humans? Without duplication possibilities, your error recovery rate will plummet, unless you build in a lot more error correction, which adds computation time and, sorry, increases the number of bits to send yet again. This is a core consideration of any distributed system, where we have to think about how many copies of something we need to send to ensure that we get one – or whether we care if we have more than one.
PLEASE LET THERE BE NO MORE!
- Oh, You Wanted Security, Integrity and Authenticity, Did You?
I’m not sure I’d want people reading my genome or mind state as it traversed across the Internet and, while we could pretend that we have a super-secret private network, security through obscurity (hiding our network or data) really doesn’t work. So, sorry to say, we’re going to have to encrypt our data to make sure that no-one else can read it but we also have to carry out integrity tests to make sure that what we sent is what we thought we sent – we don’t want to send a NICK packet and end up with a MICE packet, for simplistic example. And this is going to have to be sent down the same network as before so we’re putting more data bits down that poor beleaguered network.
Oh, and did I mention that encryption will also cost you more computational overhead? Not to mention the question of how we undertake this security because we have a basic requirement to protect all of this biodata in our system forever and eliminate the ability that someone could ever reproduce a copy of the data – because that would produce another person. (Ignore the fact that storing this much data is crazy, anyway, and that the current world networks couldn’t hold it all.)
And who holds the keys to the kingdom anyway? Lenovo recently compromised a whole heap of machines (the Superfish debacle) by putting what’s called a “self-signed root certificate” on their machines to allow an adware partner to insert ads into your viewing. This is the equivalent of selling you a house with a secret door that you don’t know about it that has a four-digit pin lock on it – it’s not secure and because you don’t know about it, you can’t fix it. Every person who worked for the teleporter company would have to be treated as a hostile entity because the value of a secretly tele-cloned person is potentially immense: from the point of view of slavery, organ harvesting, blackmail, stalking and forced labour…
But governments can get in the way, too. For example, the FREAK security flaw is a hangover from 90’s security paranoia that has never been fixed. Will governments demand in-transit inspection of certain travellers or the removal of contraband encoded elements prior to materialisation? How do you patch a hole that might have secretly removed essential proteins from the livers of every consular official of a particular country?
The security protocols and approach required for a teleporter culture could define an entire freshman seminar in maths and CS and you could still never quite have scratched the surface. But we are now wandering into the most complex areas of all.
- Ethics and Philosophy
How do we define what it means to be human? Is it the information associated with our physical state (locations, spin states and energy levels) or do we have to duplicate all of the atoms? If we can produce two different copies of the same person, the dreaded transporter accident, what does this say about the human soul? Which one is real?
How do we deal with lost packets? Are they a person? What state do they have? To whom do they belong? If we transmit to a site that is destroyed just after materialisation, can we then transmit to a safe site to restore the person or is that on shaky ground?
Do we need to develop special programming languages that make it impossible to carry out actions that would violate certain ethical or established protocols? How do we sign off on code for this? How do we test it?
Do we grant full ethical and citizenship rights to people who have been through transporters, when they are very much no longer natural born people? Does country of birth make any sense when you are recreated in the atoms of another place? Can you copy yourself legitimately? How much of yourself has to survive in order for it to claim to be you? If someone is bifurcated and ends up, barely alive, with half in one place and half in another …
There are many excellent Science Fiction works referenced in the early links and many more out there, although people are backing away from it in harder SF because it does appear to be basically impossible. But if a networking student could understand all of the issues that I’ve raised here and discuss solutions in detail, they’d basically have passed my course. And all by discussing an impossible thing.
With thanks to Sean Williams, Adelaide author, who has been discussing this a lot as he writes about teleportation from the SF perspective and inspired this post.