Would Somebody Please Keep Ben Stein from Calling Himself an Economist? Anyone? Anyone? Bueller?
A Review of Daniel Gross's Book "Pop"

Charlie Stross: Tomorrow's Future Today!

A science fiction writer as technological forecaster. I think his clients got much more than their money's worth. Whether they know what to do with it is another matter:

Charlie's Diary: Shaping the future: Good afternoon, and thank you for inviting me here today. I understand that you're expecting a talk about where the next 20 years are taking us, how far technology will go, how people will use the net, and whether big shoulder pads and food pills will be fashionable. Personally, I'm still waiting for my personal jet car — I've been waiting about fifty years now — and I mention this as a note of caution: while personal jet cars aren't obviously impossible, their non-appearance should give us some insights into how attempts to predict the future go wrong.

I'm a science fiction writer by trade, and people often think that means I spend a lot of time trying to predict possible futures. Actually, that's not the job of the SF writer at all — we're not professional futurologists, and we probably get things wrong as often as anybody else. But because we're not tied to a specific technical field we are at least supposed to keep our eyes open for surprises....

The big surprise in the 20th century — remember that personal jet car? — was the redefinition of progress that took place some time between 1950 and 1970. Before 1800, human beings didn't travel faster than a horse could gallop. The experience of travel was that it was unpleasant, slow, and usually involved a lot of exercise.... Then something odd happened; a constant that had held for all of human history — the upper limit on travel speed — turned into a variable. By 1980, the upper limit on travel speed had risen (for some lucky people on some routes) to just over Mach Two, and to just under Mach One on many other shorter routes. But from 1970 onwards, the change in the rate at which human beings travel ceased — to all intents and purposes, we aren't any faster today than we were when the Comet and Boeing 707 airliners first flew. We can plot this increase in travel speed on a graph — better still, plot the increase in maximum possible speed — and it looks quite pretty; it's a classic sigmoid curve, initially rising slowly, then with the rate of change peaking between 1920 and 1950, before tapering off again after 1970....

One side-effect of faster travel was that people traveled more. A brief google told me that in 1900, the average American traveled 210 miles per year by steam-traction railroad, and 130 miles by electric railways. Today, comparable travel figures are 16,000 miles by road and air — a fifty-fold increase in distance traveled.... We probably don't spend significantly more hours per year aboard aircraft that our 1900-period ancestors spent aboard steam trains, but at twenty times the velocity — or more — we travel much further and consume energy faster while we're doing so.

Around 1950, everyone tended to look at what the future held in terms of improvements in transportation speed. But as we know now, that wasn't where the big improvements were going to come from. The automation of information systems just weren't on the map, other than in the crudest sense — punched card sorting and collating machines and desktop calculators.

We can plot a graph of computing power against time that, prior to 1900, looks remarkably similar to the graph of maximum speed against time. Basically it's a flat line from prehistory up to the invention, in the seventeenth or eighteenth century, of the first mechanical calculating machines. It gradually rises as mechanical calculators become more sophisticated, then in the late 1930s and 1940s it starts to rise steeply. From 1960 onwards, with the transition to solid state digital electronics, it's been necessary to switch to a logarithmic scale to even keep sight of this graph. It's worth noting that the complexity of the problems we can solve with computers has not risen as rapidly as their performance would suggest to a naive bystander. This is largely because interesting problems tend to be complex, and computational complexity rarely scales linearly with the number of inputs; we haven't seen the same breakthroughs in the theory of algorithmics that we've seen in the engineering practicalities of building incrementally faster machines....

We know that Moore's Law has some way to run.... However, it looks unlikely that we'll ever be able to build circuits where the component count exceeds the number of component atoms, so I'm going to draw a line in the sand and suggest that this exponential increase in component count isn't going to go on forever.... The cultural picture in computing today therefore looks much as it did in transportation technology in the 1930s — everything tomorrow is going to be wildly faster than it is today, let alone yesterday. And this progress has been running for long enough that it's seeped into the public consciousness.... All of this is irrelevant. Because computers and microprocessors aren't the future. They're yesterday's future, and tomorrow will be about something else.

I don't expect I need to lecture you about bandwidth.... Improvements in bandwidth are something we get from improvements in travel speed or information processing; you should never underestimate the bandwidth of a pickup truck full of magnetic tapes.... Now, with little or no bandwidth, when it was expensive and scarce and modems were boxes the size of filing cabinets that could pump out a few hundred bits per second, computers weren't that interesting; they tended to be big, centralized sorting machines.... With lots of bandwidth, the picture is very different... a world where there are nearly as many mobile phones in the EU as there are people, where each mobile phone is a small computer, and where the fast 3G, UMTS phones are moving up to a megabit or so of data per second over the air — and the next-generation 4G standards are looking to move 100 mbps of data. So that's where we are now. And this picture differs from the past in a very interesting way: because lots of people are interacting with them.... It's like the difference between having an experimental test plane that can fly at 1000 km/h, and having thousands of Boeings and Airbuses that can fly at 1000 km/h and are used by millions of people every month. There will be social consequences, and you can't easily predict the consequences of the mass uptake of a technology by observing the leading-edge consequences when it first arrives.

It typically takes at least a generation before the social impact of a ubiquitous new technology becomes obvious. We are currently aware of the consequences of the switch to personal high-speed transportation — the car — and road freight distribution. It shapes our cities and towns, dictates where we live and work, and turns out to have disadvantages our ancestors were not aware of, from particulate air pollution to suburban sprawl and the decay of city centers in some countries. We tend to be less aware of the social consequences.... It is no longer rare to live a long way from relatives, workplaces, and educational institutions. Countries look much more homogeneous... because community has become delocalized from geography.... This is the effect of cheap, convenient high speed transport.

Now, we're still in the early stages of the uptake of mobile telephony, but some lessons are already becoming clear.... Mobile phones in contrast connect people, not places.... This has interesting social effects. Sometimes it's benign; you never have to wonder if someone you're meeting is lost or unable to find the venue, you never lose track of people. On the other hand, it has bad effects... bullying via mobile phone is rife in British schools.... It's even harder to predict the second-order consequences of new technologies when they start merging at the edges, and hybridizing. A modern cellphone is nothing like a late-1980s cellphone....

Putting it all together: Let's look at our notional end-point where the bandwidth and information processing revolutions are taking us — as far ahead as we can see... about 25-50 years away. Firstly, storage. I like to look at the trailing edge; how much non-volatile solid-state storage can you buy for, say, ten euros?... Today, I can pick up about 1Gb of FLASH memory in a postage stamp sized card for that much money. fast-forward a decade and that'll be 100Gb. Two decades and we'll be up to 10Tb.

10Tb is an interesting number. That's a megabit for every second in a year... enough to store a live DivX video stream... of everything I look at for a year.... It's a life log; replay it and you've got a journal file for my life.... Why would anyone want to do this?... Initially, it'll be edge cases. Police officers on duty: it'd be great to record everything they see, as evidence. Folks with early stage neurodegenerative conditions like Alzheimers: with voice tagging and some sophisticated searching, it's a memory prosthesis. Add optical character recognition on the fly for any text you look at, speech-to-text for anything you say, and it's all indexed and searchable. "What was the title of the book I looked at and wanted to remember last Thursday at 3pm?" Think of it as google for real life.

We may even end up being required to do this, by our employers or insurers.... (There are also a whole bunch of very nasty drawbacks to this technology — I'll talk about some of them later, but right now I'd just like to note that it would fundamentally change our understanding of privacy, redefine the boundary between memory and public record, and be subject to new and excitingly unpleasant forms of abuse....)

Now, this might seem as if it's generating mountains of data — but really, it isn't. There are roughly 80 million people in Germany. Let's assume they all have lifelogs. They're generating something like 10Tb of data each, 10^13 bits, per year, or 10^21 bits for the entire nation every year. 10^23 bits per century.... My model of a long term high volume data storage medium is a synthetic diamond. Carbon occurs in a variety of isotopes, and the commonest stable ones are carbon-12 and carbon-13, occurring in roughly equal abundance... a device that will create a diamond, one layer at a time, atom by atom, by stacking individual atoms — and with enough discrimination to stack carbon-12 and carbon-13, we've got a tool for writing memory diamond. Memory diamond is quite simple: at any given position in the rigid carbon lattice, a carbon-12 followed by a carbon-13 means zero, and a carbon-13 followed by a carbon-12 means one.... Sixty kilograms can store a lifelog for the entire human species for a century.... The Google cluster, as of mid-2006, was estimated to have 4 petabytes of RAM. In memory diamond, you'd need a microscope to see it. So, it's reasonable to conclude that we're not going to run out of storage any time soon.

Now, capturing the data, indexing and searching the storage, and identifying relevance — that's another matter entirely, and it's going to be one that imprint the shape of our current century on those ahead, much as the great 19th century infrastructure projects (that gave our cities paved roads and sewers and railways) define that era for us. I'd like to suggest that really fine-grained distributed processing is going to help; small processors embedded with every few hundred terabytes of storage. You want to know something, you broadcast a query: the local processors handle the problem of searching their respective chunks of the 128-bit address space, and when one of them finds something, it reports back. But this is actually boring. It's an implementation detail. What I'd like to look at is the effect this sort of project is going to have on human civilization....

[W]e're going to end up with — at the least — lifelogs, ubiquitous positioning and communication services, a civilization where every artifact more complicated than a spoon is on the internet and attentive to our moods and desires, cars that drive themselves, and a whole lot of other mind-bending consequences. All within the next two or three decades. So what can we expect of this collision between transportation, information processing, and bandwidth?

We're already living in a future nobody anticipated. We don't have personal jet cars, but we have ridiculously cheap intercontinental airline travel.... [W]e do, in fact, require more than four computers for the entire planet.... An increasing number of people don't have telephone lines any more — they rely on a radio network instead.... Hands up, anyone in the audience, who owns a slide rule? Or a set of trigonometric tables? Who's actually used them, for work, in the past year? Or decade?... [T]he pocket calculator and the computer algebra program have effectively driven those tools into obsolescence. This happened some time between the early 1970s and the late 1980s. Now we're about to see a whole bunch of similar and much weirder types of obsolescence....

[W]e'll be raising a generation of kids who don't know what it is to be lost, to not know where you are and how to get to some desired destination from wherever that is. Think about that. "Being lost" has been part of the human experience ever since our hominid ancestors were knuckle-walking around the plains of Africa. And we're going to lose it — at least, we're going to make it as unusual an experience as finding yourself out in public without your underpants. We're also in some danger of losing the concept of privacy.... [W]e're already seeing some interesting tendencies in the area of attitudes to privacy on the internet among young people, under about 25; if they've grown up with the internet they have no expectation of being able to conceal information about themselves. They seem to work on the assumption that anything that is known about them will turn up on the net sooner or later, at which point it is trivially searchable....

It'd be nice to tie your lifelog into your blog and the rest of your net presence, for your personal convenience. And at first, it'll just be the kids who do this.... Well, it'll be the kids and the folks on the Sex Offenders Register who're forced to lifelog as part of their probation terms.... Okay, it'll also be people in businesses with directors who want to exercise total control over what their employees are doing, but they don't have to work there ... yet.... The political hazards of lifelogging are, or should be, semi-obvious.... If you dig hard enough, everyone is a criminal....

And then there's history.... Barring a catastrophic universal collapse of human civilization — which I should note was widely predicted from August 1945 onward, and hasn't happened yet — we're going to be laying down memories in diamond that will outlast our bones, and our civilizations, and our languages. Sixty kilograms will handily sum up the total history of the human species, up to the year 2000. From then on... we still don't need much storage, in bulk or mass terms. There's no reason not to massively replicate it and ensure that it survives into the deep future.... [W]e're going to give future historians a chance to build an annotated, comprehensive history of the entire human race. Charting the relationships and interactions between everyone who's ever lived since the dawn of history — or at least, the dawn of the new kind of history that is about to be born this century.... I expect to live long enough to be lifelogging, but my first forty or fifty years are going to be very poorly documented, mere gigabytes of text and audio to document decades of experience. What I can be fairly sure of is that our descendants' relationship with their history is going to be very different from our own, because they will be able to see it with a level of depth and clarity that nobody has ever experienced before.

Meet your descendants. They don't know what it's like to be involuntarily lost, don't understand what we mean by the word "privacy", and will have access (sooner or later) to a historical representation of our species that defies understanding. They live in a world where history has a sharply-drawn start line, and everything they individually do or say will sooner or later be visible to everyone who comes after them, forever. They are incredibly alien to us. And, yet, these trends are emergent from the current direction of the telecommunications industry, and are likely to become visible as major cultural changes within the next ten to thirty years.

None of them require anything but a linear progression from where we are now, in a direction we're already going in. None of them take into account external technological synergies, stuff that's not obviously predictable like brain/computer interfaces, artificial intelligences, or magic wands. I've purposefully ignored discussion of nanotechnology, tissue engineering, stem cells, genomics, proteomics, the future of nuclear power, the future of environmentalism and religion, demographics, our environment, peak oil and our future energy economy, space exploration, and a host of other topics.

As projections of a near future go, the one I've presented in this talk is pretty poor. In my defense, I'd like to say that the only thing I can be sure of is that I'm probably wrong, or at least missing something as big as the internet, or antibiotics.

(I know: driverless cars. They're going to redefine our whole concept of personal autonomy. Once autonomous vehicle technology becomes sufficiently reliable, it's fairly likely that human drivers will be forbidden, except under very limited conditions. After all, human drivers are the cause of about 90% of traffic accidents: recent research shows that in about 80% of vehicle collisions the driver was distracted in the 3 seconds leading up to the incident. There's an inescapable logic to taking the most common point of failure out of the control loop — my freedom to drive should not come at the risk of life and limb to other road users, after all. But because cars have until now been marketed to us by appealing to our personal autonomy, there are going to be big social changes when we switch over to driverless vehicles.

(Once all on-road cars are driverless, the current restrictions on driving age and status of intoxication will cease to make sense. Why require a human driver to take an eight year old to school, when the eight year old can travel by themselves? Why not let drunks go home, if they're not controlling the vehicle? So the rules over who can direct a car will change. And shortly thereafter, the whole point of owning your own car — that you can drive it yourself, wherever you want — is going to be subtly undermined by the redefinition of car from an expression of independence to a glorified taxi. If I was malicious, I'd suggest that the move to autonomous vehicles will kill the personal automobile market; but instead I'll assume that people will still want to own their own four-wheeled living room, even though their relationship with it will change fundamentally. But I digress ...)

Anyway, this is the future that some of you are building. It's not the future you thought you were building, any more than the rocket designers of the 1940s would have recognized a future in which GPS-equipped hobbyists go geocaching at weekends. But it's a future that's taking shape right now, and I'd like to urge you to think hard about what kind of future you'd like your descendants — or yourselves — to live in. Engineers and programmers are the often-anonymous architects of society, and what you do now could make a huge difference to the lives of millions, even billions, of people in decades to come...

Comments