Living in a computer for fun and profit
Intro
This is based on a letter I wrote to a friend, not sure of the date (will probably dig it up at some point), probably around 2002. This is currently a rough draft and I will be gradually revising it into a form more suitable for general consumption.
Text
The idea of somehow uploading one's self to a computer is not a new one, by any means... the best example I can think of is in the later books of Fred Pohl's "Heechee" series, though I'm sure there are others. And obviously the ability to do this (specifically, the ability to separate one's awareness from one's body and allow it to continue without the existence of that body and to be able to store & transmit it as data) would be a tremendously Good Thing in a lot of ways...
But it only lately struck me just how many of my own specific problems it would solve.
- Gender dysphoria.
- Mortality
- (bonus: for the world at large, any other physical frailties which might prevent one from enjoying life very much) (starting with headaches)
- Feeling like there's nowhere safe (much less safe and comfortable) to live
- The general infeasibility of manned space travel to anywhere interesting (much less publicly-available space travel, not that I'd be able to afford it anyway)
- The inescapable feeling that society's rules and requirements are generally hostile to my needs, and yet there's no way of escaping. The only places that are unoccupied (e.g. Antarctica, the ocean, off-Earth) aren't self-sustainably habitable by humans, or at least not by me
- Being far away from friends (the "transporter" problem)
These are all Really Big Issues for me. I've tried to address each of them in various ways over the years, but the solutions are hideously slow in coming (where they've made any forward progress at all) and at the present rate I'll be lucky if any of them get even partly solved before I'm too old to appreciate it (discarding, for the moment, the notion that I'll live to be 200).
(And of course for some of these problems it may not be clear just how going to live in a computer would solve them... I'll get to that.)
#4 might seem like something that I should have overcome by now, but I haven't. There's always general stress about having to make enough money to pay for living space and food, when what I can afford space-wise always seems to feel cramped and confining (I'm spoiled by my imagination); there never even seems to be enough money that I can just relax, focus on my work, and not worry that whatever stable situation I happen to be in will remain stable.
Maybe it would have been different if I had stayed in RI and kept working at Brown, because I was doing ok financially there... or maybe if I'd stayed in Durham instead of moving to Athens, for similar reasons. Or if I'd stuck to my guns about not starting a family if the conditions weren't 100% right. Many other "if"s. Even so, I never quite felt comfortable (or safe) in RI, just stable; it's a bit of a stretch to imagine that I'd be comfortable by now even had I stayed there accumulating savings.
In other words: there was a time when I had started to gain a bit of confidence in my ability to support myself financially, and although it may be only bad luck that things ended up the way they did, but nonetheless it's all the data upon which I have to draw when determining what to expect.
#5: I've always wanted to go to space, if I hadn't mentioned this before. I used to be more obsessive about this than I am now; lately I've realized that part of it was just wanting to find a place that felt safe, where all I had to do was exist -- not go to a job or balance a bank account or anything. My equivalent of the simple life, maybe; an escape from the world of regular people, where there would just be myself and maybe a few people I felt comfortable with and could get to know without worrying that they'd up and leave in a few years. I had this subconscious idea, I think, that living in space would be like one long voyage to Jupiter -- plenty of leisure time, all basic necessities provided for, and hopefully the ship's computer wouldn't go nuts and kill everyone.
#6 has some common ground with #3:
Housing is a large chunk of one's income -- whether it's rent, taxes, or mortgage payments and taxes together. Here at Pinecrest, for example, we only have taxes to deal with, pretty much the cheapest way to go -- and yet because of the skyrocketing "property values" in this neighborhood, a house that is only just barely big enough* for the 5 of us (2 adults, 3 kids) is costing so much in taxes that we wouldn't be able to come up with it if it were not for Sandy's trust fund. Which runs out in another few years. (Never mind being able to afford the much-discussed Large House For All My Friends... although at the latter's present rate of attrition, I may have room in another decade or two because there won't be any left.) Similar houses in the neighborhood are selling for $300k+, and the taxes are based on assessed values which are based on those other sales -- even though S has no intention of ever selling (unless forced into it by bankruptcy).
(*interestingly, I'm quite sure it wouldn't be big enough at all if it weren't that so many of the things for which I used to require separate pieces of equipment can now be done on the computer.)
So, in other words, because other people with more money want to live here, we could be forced to move out unless I can find a way to make as much money as they do. If there weren't other people making that much money who wanted to live here, the property values wouldn't be going up so fast and the taxes would probably still be manageable. (This is a sort of middle-upper-class version of the "gentrification" that was going on at Fox Point in Providence when we were living there... except that the people moving in here are usually owners, not renters.)
What's ironic is that property value (at least when a neighborhood gets trendy) has relatively little to do with how pleasant the house actually is to live in, much less the quality of its construction. To actually build a house like this today would probably cost less than $100k, especially if one didn't have bureaucratic hoops to jump through and had the time to do much of the work one's self -- time one doesn't actually have, of course, because one has to be out earning the money in the first place. The value is mostly the "location, location, location". But moving the house to a less trendy location would be even more prohibitively expensive, even if one could find somewhere to put it that was both habitable and affordable.
Rants aside, even if you presumed that the house was actually intrinsically worth $300k in some way -- that this is a fair price to pay for 2000+ square feet or so of comfortable living space -- it seems to me that this is an unreasonable burden. Or maybe I mean it's too high a price for me personally; the effort it would take me to amass that much money is more effort than I have available (and shouldn't I be putting that effort into something useful, rather than amassing money just so I have a place to live?). Even if it just meant working at a reasonably pleasant programming job for 20 years, I don't know that I could do it... and lately I've had bad luck finding reasonably pleasant programming jobs, much less ones that had the potential to last that long. There are other careers I could maybe work at happily for 20 years, but I've grown suspicious of careers in general and find I no longer have the energy to invest in sideways or retrograde motion. It might well kill me to try at this point. (And then I wouldn't have the house until I was nearly 60, unless I was willing to step into that peculiar modern indentured servitude known as a "mortgage".)
The other part of #5 is kind of fragmented and I suspect I'll have difficulty coming up with good examples, although I seem to bump into them most days of the week (at which point someone generally says "oh, there's nothing you can do about that; you'd better just get used to it"). In general: society is really messed-up, and is not likely to change into anything not-messed-up anytime soon. A new society, built with new rules from the ground up, would have a chance to escape a lot of what's wrong with this one and learn from its mistakes.
For example, the way issues of public concern are discussed and debated -- it's a wonder *any* sensible decisions are ever reached; it's no wonder at all that the general beliefs about what is "right" and "moral" are based on such a morass of confusion. When people discuss something, even under the best circumstances (and the best circumstances rarely happen in the places where the decisions are actually made), the lack of consistency and thoroughness can be maddening. The question is never laid out logically, lines of reasoning are diverted and never returned to, and the entire discussion remains insanely single-threaded (most questions of any depth deserve at least a 2-d hierarchy) with the thread jumping from place to place at will. More often than not, hidden agendas are hard at work preventing too much rationality from creeping in. I could go on, but I'd probably end up writing a book (or a design spec for the debate software I've been wanting to write since about 1990).
Ok, that wasn't really an example, it was more like a pathology... an example would be the laws on marriage. Who decided that two was the exactly correct number for a marriage? (It could hardly be any fewer, I suppose, but that still leaves open the question of what's wrong with higher numbers which might be more workable in some circumstances and exactly why the choice can't be left up to the individual.) And that one must be male and the other female? And that there would be a vast body of laws which would apply when one was legally declared "married", and that these laws would apply even though none of the involved parties had seen them, much less signed a copy?
(And of course things have gereally only gotten worse lately, what with certain actions of the current U.S. administration and the resulting political climate... I was mainly speaking of things that I'd felt for a long time; recent declines only confirm and strengthen my original opinion. Our Society Is Irreparably Effed-Up, at least for the forseeable future.)
==
So now I should probably explain how I can possibly imagine that living as software instead of wetware would solve any of these problems, much less all of them. (Or is it really as obvious as it seems to me, now that I've thought of it? I have no idea.)
Let's start with what it takes to survive physically as a human, thinking in terms of effort and resources expended (which will correspond approximately with the effort required to sustain oneself).
The human body powers itself with chemical fuel. It also depends on small portions of a wide variety of other chemicals, from basic elements to highly complex organic molecules, for a variety of other purposes generally describable as "maintenance". We generally call the packages in which these various chemicals are delivered to the body "food", and it usually costs a certain amount of money -- which is generally based on how much effort and/or resources were required to produce the food in question. (Some food can be produced without spending any money, but not in any significant quantities. There are exceptions, but there are problems with the exceptions.)
The chemical fuel-burning process is not terribly efficient. (This obviously increases the cost of powering the body; more food is required.)
I don't know much about what happens with the other (non-fuel) chemicals except that the repairs and maintenance which the body is capable of doing at all are extremely limited. It cannot regrow most nerve tissue, much less a missing limb; there are countless bodily diseases for which there is no cure, or no reliable cure. Bodies (including the brain) are often mis-manufactured from the start, with no way to correct many of the defects; corrections which are possible are often very crude and unsatisfactory, not to mention hideously expensive. (Medical care is another major problem right now -- costs for most care having risen far beyond the apparent cost of the resources invested in it, much less the ability of most people to afford it.) Nor do we have any way of producing copies of the body, or parts of it, for use in the event of damage to the original.
The chemicals which the body is capable of burning and processing cannot generally be synthesized; they must be pre-processed by other organisms. This adds additional inefficiency (and therefore cost) to the production of the materials which the body uses to power itself. Many of the chemicals from which the body can obtain the most energy (e.g. meat) require no less than two sets of organisms to produce them -- e.g. a "plant" turns sunlight into fuel chemicals which humans can't burn but which certain types of "animal" can; the "animal" then eats the plant and turns that chemical energy into "meat", which contains chemical fuel(s) that humans *can* burn.
All of this requires copious amounts of open land and the ethically questionable killing of other brain-equipped organisms. (As a meat-eater I tend to avoid the whole ethical question most of the time, but I do know that I would feel a lot happier knowing that the ongoing destruction of other life-forms was not part of the chain of events leading to the creation of the energy I consume daily.)
Even if we switch to a strictly vegetable diet (which can be done, albeit with difficulty -- and with some individuals finding it much more difficult than others), we still end up with sunlight being processed twice, inefficiently, before its actual energy gets used -- at tremendous expense in resources (land, labor, topsoil, farm equipment...).
Now let's compare the costs of existing as software inside a computer. (And I suppose I'll have to get around to explaining why I think such a thing is possible... later.)
Probably the key question is, how much power would a sentience-capable computer (hmm... let's call this a Semyahost, after the computer-programmer heroine of the aforementioned "Heechee" series) require? There are too many unknowns to give a firm estimate, but we do know one thing: computers get smaller and faster over time. Computing power (resource density) doubles, and cost (including power consumption) divides in half, roughly every 18 months (Moore's Law). Going at this exponential rate we will eventually reach just about any needed level of complexity and power consumption within a relatively short time (and the time at which it will be reached can be predicted with some accuracy, once we figure out what it is).
Let me try and pin down some practicalities, if only to help paint a picture of where Semyahosts might be seated and what they might look like. Let's suppose that a Semyahost (large enough to hold the equivalent of at least one human consciousness -- which I'll call an Essienaut -- and any other personal data the Essienaut needs in order to exist happily and productively) is about the size of a large network server of 2003: maybe half a cubic meter in volume and using maybe 500 watts of power. (For reference: I believe such a machine today would have about a GB of RAM, a few hundred GB of drive space, and several TB of online removable storage.)
A Essienaut would not need physical "living space" except for maintenance purposes; its living space would be *inside* the Semyahost. If this sounds dreadfully dry and utopian (or else just baffling), then I guess I'd better explain (later) why I think it would work that way...
But anyway -- not needing elbow room, an Essienaut could live in (a Semyahost in) someone's basement or warehouse (connected with others of its kind on the local network -- possibly sharing one or more CPUs between them, depending on just how much CPU power is needed per Essienaut and how much is available per CPU) and presumably to the Internet (or its successor) and perhaps some external sensors and actuators, depending on the individual Essienaut's "job" and interests), using maybe $300-$400 worth of electricity per month at 10 cents/kwh (which is a high estimate; it's about 7.5 here).
That being more or less equivalent to the cost of rent (albeit cheaper than most habitable apartments in the country right now), you might find yourself thinking "gee, why bother doing this if it's still going to cost that much to live?". You have to remember, though, that this is the Essienaut's equivalent of food costs; her living expenses, being in an un-heated part of the house and not needing much space (or other amenities, e.g. plumbing, clothing, toiletries), are next to nothing. Semyahosts could be kept in a plastic storage shed in the back yard. They just need to be kept away from moisture and extremes of heat, safe from theft or damage, and accessible for physical maintenance (more about that later).
That's if they live in places where humans also live, of course. The minimal physical requirements of the Essienaut lead to another fascinating possibility (which is probably required for solving some of the other problems I mentioned at the beginning).
Regardless of how much room, materials, or energy an Essienaut may turn out to require, she has two huge advantages over a human: (1) she doesn't need air, and (2) she can be fed entirely by energy converted from sunlight at an efficiency of approximately 40% (given present technology) via solar cells. Off Earth, the equation changes completely; an Essienaut is no longer concerned about the cost of the utility company electricity needed to power herself, because there is free power available for the collecting. Her main expense will be getting her Semyahost (and its solar cells) into space, and keeping it operational there. Any given human wishing to live in one (to become an Essienaut, in other words) need only be digitized and transmitted, once a Semyahost is available; no physical proximity is needed.
[ NOTE: This is absolutely maddening... I just read, within the past few weeks, a science article which offhandedly mentioned how much energy the sun delivers to Earth's equator at noon, per square meter, among other useful figures... and I can't remember the figures nor where I saw them. I'm going to be conservative and guess that it was about 25 watts, though I think it was higher -- and would almost certainly be higher in space. 40% of that is about 10 watts. ]
Assuming we can draw 10 watts/m^2 in sunlight anywhere near Earth's orbit, we need 50 m^2 (just over 7 m on a side) to power our Semyahost -- maybe $10k worth of solar cells? (Certainly minor compared to the cost of launch, which would mean it's probably the cost of launch that's going to be the bottleneck.) 50 m^2 is no big deal in interplanetary space, and hardly even worth worrying about on some place like the Moon. For a community of any size, though, we'll need a lot of solar cells... the cells don't have to be terribly close to the Semyahosts, though; they could use PowerSat technology developed in the 1970s (and no doubt refined since then), but with fewer risks and greater energy efficiency. Plus they could be maintained by robots remote-piloted by members the local Essienaut community... but I'm getting ahead of myself.
It seems to me that space would be the natural habitat for Essienauts -- plenty of room and energy for their Semyahosts. I imagine the first Semyahosts would be built on Earth and rocketed into space; perhaps later Essienauts would build mines and factories (note that Luna is a rich source of many materials used in semiconductor manufacture) and become self-maintaining. I imagine a small trade developing between the spacedwelling Essienaut community and certain Earth businesses -- think of a company (or government) having to spend millions launching a new satellite to replace old one, and then imagine how much they'd be willing to pay a Semyanaut-operated remote to jet over and replace a failing component or three; certainly there's money to be made in space. The problem has always been the prohibitive cost of sustaining the humans who went there.
Fuel for maneuvering... ok, that's something we can't just manufacture from raw solar power... perhaps the solar wind could be harvested? Technological advances aside, at first we'd need fuel shipped up from Earth; this is still orders of magnitude less expensive than shipping up fuel AND food AND air AND fresh personnel... (Humans are very high-maintenance.) Barring exotic solutions, there's probably some way to make fuel out of stuff found on the moon (and use a mass-driver to get it into orbit, instead of using up the fuel)... or the asteroids, or somewhere *not* buried at the bottom of an atmosphere-choked gravity well. (Actually, as long as we can find some way to use electricity to continuously expel mass with reasonable efficiency, all we *really* need is reaction mass, which can be almost anything.)
Carl Sagan, otherwise a great advocate of space exploration, always maintained that sending humans into space was excessively risky and expensive when we could send probes to do the same thing (or better, actually). I used to shake my head in despair at this attitude; he just didn't *get* it... With electronically-hosted intelligence, however, we can have the best of both worlds -- being "there" every bit as much as a human would be and perhaps more (digital video cameras now rival the best the eye can do, and of course human visual perception is limited to an extremely narrow frequency band), at essentially no personal risk and requiring no more logistical support than a probe would require. We can explore -- even live in -- places that would be lethal to a human. With the possibility of backups in a safe location, we can even personally explore places dangerous to a robotic probe, with only the risk of losing the memory of the exploration (or part of that memory; it should be a fairly simple matter to keep the "master" Essienaut updated with a live feed of experiences and reactions from the remote copy -- though bandwidth may be a problem...)
==
Remaining issues to discuss: - why I think it's possible to become software - why I think an Essienaut won't need living space - more about remote-controlled Robots
These seem best tackled in reverse order.
About Those Robots
It seems to me that a lot of our current ideas about What Robots Can't Do have more to do with the limitations on our ability to control them (or for them to behave intelligently on their own) than with any real limitation in the mechanics of the unit itself (though such limitations certainly do exist). Robots have difficulty climbing stairs, driving cars, making sensible decisions about what to do next (e.g. when exploring distant regions where decisions from Earth can take minutes or hours to arrive), manipulating delicate objects, and so on...
With an electronic person controlling a high-quality robotic unit, that unit should have most of the abilities of a regular human person (and some extras that a human wouldn't have). Video and audio sensors today have better resolution than the human eye and ear (I think); the main deficiency in robotics, as far as I know, is touch. (Electronic odor sensors are also quite primitive, but this is fairly unimportant for most day-to-day tasks -- especially if cooking and eating aren't among them.) One of the main problems with developing better sensors, of course, is that there's no way to hook them up to the appropriate areas of the brain for signal processing; this will not be a problem for Essienauts, and I imagine progress in cybernetic sensing will be dramatically accelerated by both demand for improvements and the ability to test those sensors by direct hook-up.
Living space
Given what it will be possible to simulate in 30 years versus what can be built in the physical world (and how much it costs and how long it takes and how difficult it is even to get started), I can't imagine why anyone would want to make their home anywhere except in virtual reality.
[ ASIDE about "reality" ]
Even what it's possible to simulate today can be something I wouldn't mind seeing out of my living room window. (See, for example, [ http://digitalblasphemy.com ].) Photorealistic images such as these currently can take hours or days to render, so giving them motion and life is a bit of a problem -- but I can only think that sometime before we reach the level of computing power needed in a Semyahost, we will have enough to simulate quite realistic scenery in real time. Recent experiments in "artificial organisms" which simulate realistic patterns of growth will, I suspect, cross-fertilize with image rendering technology and eventually evolve into entire simulated landscapes.
If the simulation is detailed enough, who is to say whether or not it is "real"? What if we are, for example, simulating down to the atomic level, such that the unusual chemicals produced by a particular strain of artificial life can be tested for their effect on artificial human bodies? What if a hitherto-unknown artificial plant found growing in a simulated rainforest turns out to produce chemicals with beneficial medicinal properties? Eventually there must arise some way to reproduce those chemicals in "reality", either by recreating the genes that make the plant or by synthesizing the chemical.
The ultimate test, of course, will be when items in virtual reality can reliably be created in "reality". We can do this pretty well with sounds and (at least slowly) with flat motion video; life forms are a bit more tricky, but there's no reason to think it won't eventually be possible. At that point, "virtual reality" becomes just another place in "reality", with the difference being that we are able to manipulate the laws of that "reality" at will. With another few magnitudes (beyond the Semyahost-level) of increase in computing power, we may eventually find ourselves preserving entire ecosystems digitally.)
[ end of ASIDE ]
One might even wonder why people would bother with such things as houses and scenery when living in cyberspace. Let's break this question down a bit: what function would these fixtures serve in a virtual environment, where physical laws exist only at the whim of the occupant?
Here's what I think the answer is: houses and scenery each serve a useful function, even in the abstract.
A "real", physical house, especially the more function-specific parts of it (the office, the machine shop, living room, the bedroom...) serves as a kind of graphical user interface (UI) for the things we do in those rooms. If you keep some good bedtime books in a shelf over your "bed", or put something down on a "table", you know those things will be there later (unless moved by other, perhaps unwanted, elements of reality -- children, pets, another person, a breeze). You put things in your bedroom that you expect to want to use during the process of getting ready for bed, sleeping, or waking up (and as few unrelated things as possible); "walking" into the "bedroom" sets the context for what you will be doing (in this case, sleeping) -- a little bit like navigating into a particular project folder on your hard drive. Same for the machine shop, the music room, the library.
"Real" scenery serves as a kind of low-intensity entertainment. (We're talking
- enjoyable* scenery here; city-dwellers can imagine a panoramic picture-window
view from the 98th floor, and country-folk can imagine looking down from a mountainside, or into a forest, or whatever. (joke -> ) I'm not sure what suburbanites should do; perhaps a view of the golf club or the mall...)
To put it in terms familiar to contemporary computer users: a house is the graphical interface, and the scenery is the "wallpaper". We have already found it useful to model certain behaviors of a physical desktop in the computer's graphical environment (the graphical "desktop", with its "windows" which are actually more like pieces of paper or perhaps wall-less resizable stackable trays whose contents can't accidentally be spilled). The difference is that when the reality of the computer surrounds you instead of only occupying a flat square space approximately one foot square, the components of the user interface can become three-dimensional and tangible. You won't "click and drag"; you'll "lift and carry", albeit without risking back injury or pinched fingers. As we become more familiar with these environments and more adept at devising them, they will no doubt become more realistically detailed -- in much the same way as many applications and OSs today allow "skins" and "themes" which make the user interface components seem more like physical objects. The richness of detail, properly used, gives us almost-subconscious cues which allow us to make better use of the components.
Becoming Software
The ability to "upload" oneself depends on two things, both of them difficult. First, you have to have a computer that is powerful enough to replicate the consciousness functions of a human brain. Second, you have to have some way to "copy" the "data" in an existing brain into that computer. The second part is a real bear, so I'm concentrating on the first part (one problem at a time). If we can create *native* Essienauts, i.e. not copied from existing humans but born and raised in Essiespace, perhaps they will have sufficient clarity of thought to help us with the second problem... or maybe someone else will solve the second problem while I'm working on the first one.
"The human brain contains somewhere around 100 billion neurons and 100 trillion synapses." [ http://www.singinst.org/seedAI/general.html ]
Oversimplistically modeling a synapse as a single-precision floating-point number (4 bytes), this means we'd need 400,000,000,000,000 bytes (400 terabytes) just to store the data contained in a typical human brain.
A good portion of those 100 terasynapses are probably in the autonomic sections of the brain and may be unnecessary (when an Essienaut is not using a robotic extension, at any rate); we may also get lucky and find that 4 bytes of resolution is unnecessary for many synapses. On the other hand, we will probably need additional storage for the Essienaut's work and daily life (think of all the things you "store" in your immediate environment)... so 400 TB seems like a reasonable guess of what we'd need in order to get started, if not to live in digital luxury.
Starting with a present low-budget disc storage capacity of 120 GB (about $100 at today's prices) and doubling every 1.5 months, how long does it take to reach 400 TB? Approximately 12 doublings gets us from 120 to 400,000 -- 17.5 years. That's NOT LONG.
Ok, let's be pessimistic... we need 400 TB of RAM, not disc; disc space is too slow. Present RAM sizes generally start around 128MB, but it's not hard to find system boards supporting over a gigabyte (and it's not terribly expensive to fill them up). So let's start at 1 GB. 18.6 doublings gets us from 1 to 400,000 -- not quite 30 years.
Now, CPU power...
"Neurons are slow; a neuron has a firing rate of only 200 times per second, and the signals travel at a maximum speed of only 150 meters per second." - ibid.
One might think "oh, THAT one's EASY!" but you have to remember that this is EACH of those 100 terasynapses firing 200 times a second (ok, processing signals from neurons firing at about 200 times/second -- same thing, computationally), all working in parallel... or in other words, (100e+9 x 200 = ) 20,000 teraflops/second. Determining flops for today's microprocessors is tricky because it really depends what sort of processing you're doing, but here's my best estimate. AMD's latest 64-bit processor (available at retail for about $500) can be clocked at up to 2,600 MHz and can run two 64-bit (8-byte) operations per cycle. Boldly extrapolating that this means it could run four 4-byte floating point ops in those two cycles, where there are 2,600,000 cycles per second, that means 10.4 gflops.
From 10.4 gflops to 20,000 tflops (20,000,000 gflops) is... almost 21 doublings, or 31 years. (And who's to say that we won't come up with some specialized hardware for this stuff before then?)
So it seems clear to me that barring major setbacks, we should have quite adequate computing power (at least to set up a rough digital homestead) by the year 2035 or so.
The real problem (the "second" problem alluded to above), of course, is doing the readout of the human brain. A complicating factor is the apparent question of just how do you avoid merely *duplicating* yourself? I mean, you attach the uploading cable to the side of your head, you press the "upload" button on the Windows 2040 screen (sorry, no, it'll be 1024-bit Linux, not Windows; I have no wish to face the Blue Reaper in my quest for immortality) and... there you are, onscreen, waving back and reporting a successful journey. But wait, you cry, I'm still here!
I think we'll have to understand a lot more about how consciousness works before we can really solve this problem; the obvious way to do it complicates the transfer process (which is already enough of a problem) immensely: As each synapse's value is copied over, we have to somehow take that synapse "out of the circuit" and let the digital copy take over. I don't even want to *think* about the bandwidth required to do this... Hopefully, consciousness theory will show us an easier way. (If I understand correctly, only a very small portion of the brain's umpty-zillion synapses are actually involved with consciousness, and it may turn out that only a small portion of those are key players; this suggests some possible avenues of inquiry, once we get anywhere near that territory.)
Back to the readout problem. (What, me avoid a difficult subject?) The Singularity Institute seems to be convinced that we'll need nanotechnology, and perhaps we will. I can only think there has to be a quicker way. It used to be thought that we could never know the chemical makeup of a star; it used to be thought that we could never see inside the body without surgery, or see what areas of the brain were being used most heavily at a particular time. Along came stellar spectroscopy, and x-rays, and CAT scans, and magnetic resonance imaging (MRI)... Perhaps some complex hybrid of MRI, or something like it, and signal analysis from a very high-resolution web of probes on the surface of the skull could get us enough information. Perhaps not-quite-nanotech will allow us to put computers into capillary-sized packages, and we can get our readouts that way.
Or perhaps nanotech will be around by 2040. Either way, we have a lot of other problems to solve before we'd be ready for uploading; no reason not to go ahead and work on them so they'll be ready and refined when the readout technology hoves into view.
As for what I optimistically fantasize I could do... I don't know precisely, but having worked with neural networks a bit and thought a great deal about what it means to "think", I've got some ideas... and what I think is that the details are probably best left for a separate chunk of writing because it's taken me way too long to get this one from 99% complete to actually sending it.