The following are some informal projections of possible technology-oriented developments that I made back in late 1999 to early 2000, while I was working on part two of "Accelerando" ("Troubadour"). This was at the time when I was trying to figure out where we're going for purposes of writing a near-future SF novel. I figured at the time that they were probably wrong, but so what? You've got to start somewhere if you're going to take a stab at designing the future.
Today they look curiously dated; short term political crises have overrun them, stealing the headlines, and I missed out the urgent proximity of Peak Oil (although I still paid some attention to it.). Still, they're interesting to look at by way of an anchor point for what I was thinking at the time I was writing the book. These days I'd add some technologies (Machinima, Cognitive Radio) while probably dropping others. But this is a document of its time ...
Firstly, that the apparent state of the economy (anywhere in the developed world) is a superposition of three different growth curves -- the old economy (shipbuilding, steel, textiles -- in steep decline), the service economy (everything from restaurant waiters to bus drivers -- expanding in proportion to the overall size of the economy, i.e. slowly), and the new economy (information tech, engineering, biotech, aerospace), growing very rapidly indeed.
The GDP growth we see (4% in the USA, 2-3% in the UK and EU, up to 8% in parts of the Pacific Rim that have managed to recover from the early 90's) is a combination of these three sectors -- take the hi-tech growth rate, add the service sector growth rate, subtract the old economy. Hmm. Rather than talking about growth rates, this should be growth rate multiplied by size of the economic sector in question. So if you've got a small but rapidly growing high tech sector, and a huge but slowly shrinking old tech sector, the overall curve will be either slow shrinkage or very slow growth -- until the old tech sector vanishes completely or hits baseline, at which point suddenly the economy will appear to surge as the high-tech sector growth rate becomes significant.
(To put these growth rates in perspective, the UK high tech sector grew steadily at an average of 20% per annum from 1990 through 1996, IIRC; it's now believed to be entering a steep recession, in which it is expected to grow at only 11% per annum! Compare _that_ to traditional economic growth rates. NB: Look here for a fun view of SF Economics and super-fast economic growth.
So. If you're a coal miner or blacksmith or farmer the future looks bleak; if you work in MacDonald's it's business as usual: and if you work in biotech or nanotech or software it's so bright you've got to wear mirrorshades. What else can we predict, trivially?
Moore's Law -- that the power of microprocessors will double every 18 months -- is looking a bit hairy these days. Firstly, the doubling time has contracted to every 15 months. Secondly, the limit to shrinking track sizes (below which quantum effects like tunneling screw our existing technologies over) is within grasp -- it's only another 5-10 years away. However, once we hit the minimum scale for today's architectures, chips can carry on growing for a bit longer. First, they can go 3D, with multiple slabs of circuitry layered on top of each other (subject to solving the heat dissipation problems). Secondly, quantum computers were science fiction in 1991; in 1994 some theoreticians came up with algorithms for exploiting these SF-nal devices; in 1996 somebody was muttering about building a "qu-bit" storage device; then in 1999 the NSA suddenly shut up about banning exports of strong cryptography tools -- strong, if you assume that finding the common prime factors of a long number is an intractable problem. If you have a working quantum computer it's anything but, and the NSA are rumoured to be 5-10 years ahead of the commercial state of the art in supercomputing ...
Even leaving aside spook paranoia, I'm inclined to assume that the sheer momentum of the hardware industry will force them to develop and productize quantum computing technology in the 7-15 year time frame -- because if they fail to do so, they will enter a massive recession, with growth rates dropping towards zero as the market is saturated by existing technology.
Predictions for 2001 show growth in PC sales volume in the USA flattening; anyone who has any use for the blessed things is already on the treadmill of buying and regularly upgrading already, so there's no additional latent consumer base to tap for growth. PC prices have fallen by 50% over the 1990's; if this trend continues, in 2010 a PC system (with 10 GFlops of processing capacity, 16-64Gb of RAM, and >1Tb of storage) will cost US $500, and in 2020 it will cost $250 (while giving you 1Tb of memory, petabyte-range disk storage sufficient for every Hollywood movie ever released, and processing power about on a par with a rat). But taking into account increases in GDP, that 2020 price of $250 will take as big a bite out of your income as $50-100 today. In other words, students and the unemployed will be able to buy new PC's. In contrast, a PDA equivalent to today's Palm IIIc will, by 2020, have more processing power and storage than the 2010 PC, and cost the equivalent in today's terms of $10-30. They'll be viable in the third world, and be well on the way to being a cereal-packet-coupon item in the west.My guess is that as 3G services are rolled out during the 2002-2008 time frame, they will converge with handheld computing (PDA's -- duh), ubiquitous streaming media (music and video), and internet access -- they'll also render the current model of telecoms development in developing nations obsolete. (No more fixed lines to villages; instead, everyone will use cellphones and packet radio systems). They'll also have unpredictable effects on the future of education: ubiquitous data networking provides the infrastructure for ubiquitous distance learning. If you can get kids in outlying remote areas up to basic literacy, then drop a communications widget (that's so cheap it's effectively free) on them through which they can participate, you've eliminated some of the constraints imposed by isolation. Ubiquitous distance learning will have implications that aren't immediately obvious; for example, consider what happens when every [currently illiterate] village lad in Afghanistan can study with the best Islamic scholars of the age via cellphone? Speaking of the developed world, expect the university degree to become obsolescent. In the high-growth new tech sector, a degree is obsolete within a decade already; if it takes three or four years to study for, then the odds are that the syllabus you sign up to study at 18 will cover nothing but ancient history by the time you graduate. The devaluation of degrees isn't going to be evenly applied to the whole of academia, however -- I'd expect a liberal arts grounding to be a lot more future-proof than an engineering or biotech discipline. Again, this doesn't mean that nobody will do scientific or technical degrees: merely that they'll have to focus more on tools for critical thinking and analysis than on specific technologies, because the technologies will become increasingly evanescent.
Now for a major neck-sticking-out prediction: four laws that will govern the next couple of decades:
- Anything that can be mass-produced will be commoditized, and the price of such goods will decline until it converges with the cost of the raw material required to fabricate them,
- Intellectual property is an oxymoron,
- The definition of "poverty" is a moveable feast.
- The means of production are also a moveable feast.
Of these assertions, I figure that #1 is relatively non-controversial. If you want an example, think of the humble VCR; back in 1980, a VCR would cost roughly a thousand pounds. Today the price of a new one, considerably more functional than the 1980 luxury edition, is down to 70 pounds or thereabouts, a large chunk of which is tax; inability to afford a colour television (itself a luxury item when introduced in 1966) is now officially one of the defining characteristics of poverty in the UK.
If you want more support for the commodification idea, I urge you to read two long items that are on the web: Unbounding the Future (by Eric Drexler) and Philosophy of the GNU Project (by Richard Stallman).
More on the latter in a moment ...
Assertion #2 is more provocative (not to say controversial). In a nutshell, part of the reason for the controversy comes from confusion over the two meanings of the word "free" in English (free as in speech, versus free as in beer). Another part of the problem lies with our collective amnesia over the reason why intellectual property rights were first promulgated -- for these rights are constructed, rather than natural. (I've written about this in my Computer Shopper column.)
One point that isn't made explicitly by Drexler or other proponents of nanotechnology is that a nanotech assembler -- a molecular robot capable of assembling other arbitrary constructions (including other assemblers) out of atoms, in accordance with some sort of instruction system, is a compiler. A compiler, in CS terms, is a piece of software that takes input in one language and translates it, producing output in a different language (typically machine code). Given that there's a uniform mapping from the instruction code for an assembler to the output format (complex molecules), we can consider the logical effects of a 'mature' nanotechnology (one that gives us general-purpose assemblers) will be to apply the economics of software production to mass-produced goods. Premium designs will be sold at premium prices, and regular upgrades marketed heavily at consumers; but at the bottom end, groups like the Free Software Foundation will generate core categories of design for products that can be assembled by replicator -- and the cost of owning such products will tend towards the cost of energy and raw materials that go into them. A bicycle, in real terms, will cost one euro or one dollar; a bicycle, if you want the latest in exotic design by a specialist company with a hot rep, will cost exactly as much as it does today as a proportion of your income.
Technologies that will not become de-commoditized are those that are mediated via a service industry. Medicine, for example: nanotech may make tissue repair easy and synthesis of exotic drugs or genome treatments feasible, but it's still going to need to be mediated by personnel with extensive training (and probably expensive professional liability insurance). So commoditization won't bring healthcare costs down -- it'll just improve the likelihood of a successful outcome for any given case. The real impact will be on chronic care: if a lot of chronic degenerative conditions (such as Alzheimer's, or the complex of syndromes collectively known as "old age") become treatable, we will see shrinkage in the medical sector and big social upheavals (although not within the time scale I'm talking about).
The third and fourth points hinge on the definition of poverty. We all have certain things without which we consider ourselves to be impoverished. First on the list are drinking water, food, and shelter: without any one of these, we may well die. They're followed by clothing, education, and access to transport -- we can live without these, but our lot would be pretty dismal. Next up the ladder we come to former luxuries: indoor plumbing, television, radio, washing machines. These were all once luxury or premium items, but have now been commoditized until lack of them is an indicator of poverty. We can expect other luxury items to join them in the pool of things that even the poverty-stricken can expect to have: internet access, ubiquitous computing, gourmet foods, foreign holidays. Poverty in the west today is sometimes defined in terms of social exclusion -- you're poverty-stricken if you can't participate in the activities of society at large -- and this definition will harden because the poor, for all their useful widgets that come courtesy of designs copylefted by the Free Hardware Foundation, won't be able to afford the latest designer labels or the energy to travel extensively.
Because a mature nanotechnology permits production anywhere, all physical goods become a bit like software today. Software can be produced by geeks in India or Indiana, and the former may cost a tenth as much to employ as the latter. The migration of jobs away from the "rust belt" industrial zones between the 1970's and 1990's will be recapitulated frequently and with bewildering speed and completeness in the 21st century. Silicon Valley will go back to orange farming in a matter of months when the silicon becomes obsolete; new industries will go where the skills are, and the skills will be as much a matter of attitude and education as much as anything else.
Did I say silicon? Silicon's already dead, man. Intel, Motorolla, they're dinosaurs. The real action is in organic semiconductors -- semiconductor inks printed onto sheets of paper or plastic using the hypertrophied descendants of today's inkjet printers. They aren't faster or bigger or better: they're cheaper, to the point of disposability. Today's disposable paper cellphone will be followed by tomorrow's disposable laptop, by supercomputers that come in paperback bindings and sit on library shelves, chattering to one another by ultraviolet laser light.
But obsessing about computers is a very 1990's sort of thing to do -- by 2010 it'll be like obsessing about steam locomotives. The real action will have gone elsewhere.
Energy is something to keep an eye on. The price of petrochemicals isn't going to come down; it's going to go up, savagely. But this doesn't mean you'll have to give up your travel habit. We're already seeing hybrid engined cars on sale, and fuel cells for home electricity generation go on sale next year. Some specialists are predicting the demise of the central electricity-generating power station within the next twenty years, as homes move over to generating their own electricity from natural gas using high-efficiency fuel cells. This makes for an overall efficiency gain, especially when we consider transmission losses in the grid. Whatever the long-term oil reserves are like, we aren't going to run out of methane sources in the next thirty (or sixty, or hundred) years (although there's trouble ahead -- the EU's biggest source of natural gas is in Russia, which has interesting political implications from 2020 onwards). Widespread use of fuel cells means we end up with electric automobiles as fast as today's IC models, with superior energy efficiency and as easy to refuel as a petrol-guzzler.
Real estate isn't impacted by the technological change dynamics (much). People with more posessions want more space to store them. There's also a trend towards one-person households in the west as a whole. Population growth in Europe has dropped below replacement levels: in the USA it's propped up by immigration, but ZPG isn't too far away. So I'd tentatively expect house prices to stabilize and maybe even decline in the long term -- except in places where people want to congregate, where scarcity will drive competition. (We have such places already: they're called "cities".) I wish I could point to a future where Buckminster Fuller's ideas about houses and technology would take off, but I can't, at least in the short term -- although camping equipment in 2030 will look very exotic to our eyes. (Self-assembling geodesic tents with their own mains-voltage electricity supply driven by fuel cells, self-inflating furniture, and air-cell insulation. A home big enough for a family of four that packs down into something the size and shape of a day-sac and weighs five kilograms.)
Entertainment ... leaving aside the copyright wars, the main studio screen stars will face stiff competition from Charlie Chaplin, Marilyn Monroe, and all the dead stars of the past. Amateurs will be making movies with the SFX and production values of 2001 Hollywood big-budget productions in their back office by 2020. Possibly we'll have seen the decay of our IP infrastructure kill the big studios by 2030; then again, maybe they'll find a way of fighting back.
A huge political storm will brew up over copyright and the concept of intellectual property in general. This is already beginning (the Napster/p2p controversy is just the gust front); on the one hand we have huge old media corporations, and on the other hand we have a consortium of consumers and producers (authors, artists and musicians) who are currently exploited by those old-media corporations. Expect draconian copyright enforcement laws ("own $1000 retail value in pirate software, go to prison for five years"), an increasing migration to free software and media distribution formats (e.g. new bands releasing promo material on MP3 or Ogg Vorbis), and new business models to take advantage of the new media (e.g. the Linux distributors, who went public selling something that's free).
The trend among IP corporations towards charging for renting media and software will accelerate (as witness Microsoft's .NET, the IT industry's dash towards ASP, the RIAA's lawsuit against Napster and attempts to charge per download for music, and the small print in the phase-two SDMI specification). These are attempts by the buggy-whip manufacturers to buck the trend and make horseless carriages illegal; in the long run they're hopeless, but in the short term they're messy and will inconvenience a lot of people.
This brings me as far out as I feel like writing right now -- this is simply an attempt to scope out the period 2000-2030, without going unduly overboard on the implications of mature nanotechnology, development of artificial intelligence, mind uploading, or the other gosh-wow trappings of extropian SF. These are predictions I hope to be around to feel embarrassed about when the deadlines roll round ... the rest will come later.
Finally, expect one entirely new magical technology to come out of nowhere and sandbag everyone who wasn't watching the ball roughly every five years (1990's), three years (2000's), and then annually or faster (2010's).
The 1990's examples were the internet and genetic engineering. The 2000's are shaping up to be quantum computing, stem cell derived tissue regeneration, microtechnology, and maybe Bose-Einstein condensate manipulation ("atomic holography"). The 2010's will include mature molecular nanotechnology and cthulhu-only-knows-what-else. Sapient business models? Practical applications of Higgs bosons (e.g. for producing new states of condensed matter)? Mind uploading/AI/EI?
The jury is out ...