Stross seminar

1. Stross on development economics (Krugman)

Civilizations are complicated.

That statement ought to be ploddingly obvious to the point of banality, but it’s astonishing how often it seems to elude pundits, politicians, and — yes — science fiction authors.

As Paul Krugman observes, we don’t really know why development economics started working better around 1980. I’d go further: I’m not sure 1980 wasn’t simply a coincidence. All we know for sure is that given access to a sufficiency of tools and ideas, sometimes a nation or group of nations (or a region within a nation — huge parts of China’s interior still remain locked in peasant farming poverty) figures out how to build institutions and infrastructure at a dizzying rate, only slowing when they near the then-prevailing state of the art. (Which itself is moving forward only slowly.)

The Merchant Princes series is to some extent a failed thought experiment in development economics. (I say failed because, for various reasons, the series is probably ending with book six; the scale of the canvas exceeds my ability to do it justice, and the style of the series — effectively a series novel, where each book is a chapter rather than a stand-alone — makes it difficult for me to remember what I’m doing.)

But it’s also shaping up as a morality play about the dangers of blithely walking into a situation and attempting to impose reorganization from the top down.

Miriam’s intervention in Clan politics in the second book (The Hidden Family) generates blowback, with a vengeance, because she’s failed to realize that the changes she is proposing will destroy the power base of a group of elderly women who, through their iron grip on the arranged marriage structure of the Clan, have carved out a tolerable niche for themselves in an otherwise intolerable world. She’s challenging the business model that has made the Clan’s conservative faction wealthy (and as we know, the first rule of politics in any place and time is “don’t be disrespecting the Money”). And she’s provoked them into actions that result in counter-actions outside the Clan, by antagonizing the monarchy and indirectly exposing the Clan’s existence to the US government. Societies, as I noted earlier, are complex: there’s never just one power center, no matter how centralised a culture might appear to an outsider.

It’s all a house of cards, a nest of delicate interlocking dependencies. Trying to introduce change is one way to kick-start the development process; but too many changes, too fast risks generating revolution or civil war, not to mention massive disenfranchisement and deprivation among the general public (as suggested by Klein’s “The Shock Doctrine”).

2. Money makes singularity (Quiggin)

Here’s a confession you won’t hear too often: I’m ignorant and under-educated — especially in economics and finance (but I’ll cop to the arts and languages too, if push comes to shove).

Trying to learn about somebody else’s discipline when you’re an outsider is an interesting experience. The first stage is bafflement, as you’re confronted by a thorny hedge of impenetrable jargon and recursive definitions. The second stage is over-simplification, as, equipped with a Bluffer’s Guide level of understanding of some of the basics, you pry apart the thorny branches and decide that the jargon is, in fact, a Shavian conspiracy against the laity and conceals the essential simplicity of the field in question. And the third stage in learning occurs when you push further into the hedge, and the branches behind you whip back into position behind and impale you on the thorns of your own misconceptions.

I’m a bit like that with money. (It’s probably why I write for a living, rather than being a hedge fund manager.) Because, when you get down to it, I don’t understand money. In fact, I’m not sure anybody does. So I’m going to retreat towards more solid ground and talk about an area where I’m at least able to grasp the scale of my own ignorance: programming.

There’s a paper I read a couple of years ago — I ran across it on the internet, but can’t find it right now — by a couple of eminent computer science academics, discussing the failure of the first fifty years of teaching programming. Fifty years ago, they explained, they could take a class of students and by the end of the class approximately 50% of them would master three essential abstract concepts. (In ascending order of abstraction these are: named variables, loop constructs, and pointers.) The other 50% of the students would flail around, programming by cutting and pasting chunks of code from elsewhere, without real understanding — as could be demonstrated by testing. In the early years of the 21st century, the outcomes are no better: 50% of folks who try to understand programming simply don’t seem to be able to grasp the core abstractions, especially pointer indirection. (It’s an inability I can sympathise with: it took me ages to get my head around what was going on.)

Money, it seems to me, is an indirection layer between barter transactions — a pointer that can reference any number of types of variable (or commodity). You can do arithmetic on money, establish how many oranges are equivalent to a gross of apples, and convert between types! (But don’t be surprised if your conversion of gasoline into lemons fails to fill the fuel tank of your car.) It lets you encapsulate a whole lot of information in a single unidimensional variable. And then …

You hit the third stage of enlightenment and the bramble patch bites you on the ass.

Money is a unidimensional signal: it tells us how much a participant in a transaction is willing to pay for something, but not why. It gives no measure of the internal state of the participant. Direct barter is more obviously amenable to theory of mind, to the participants gaming each others’ inner states — but it works really badly when you’re trying to keep a complex supply chain running. And I can’t help feeling that the unidimensional nature of the information encoded by money is somehow responsible for many of the problems we’ve seen over the past year. Let’s take a random example: Would we have had a housing bubble if houses — real estate bought, used, and sold over a period of decades — was denominated in, call it, “slow” money, accounted for and evaluated over many years and only used in the housing market, which could not be interchanged directly with our everyday “fast” money (profit and loss statements due quarterly, please), used in every other transaction? A second type of money — or, from another angle, money that encodes a different type of information — might have kept the damage from propagating. (Although I’m inclined to think that some brilliant financial super-programmer somewhere would have figured out a way to leverage the slow money in the housing market to build fast money futures, as a way around the barrier. Idiots are ingenious.)

Where was I? Oh: Accelerando.

I began writing Accelerando in 1999, in the middle of the first dot-com bubble. I’d been hired on contract in 1997 to be the first programmer in a start-up. We were writing software glue to allow merchants to accept credit card payments over the internet. Back in 1996, nobody was doing this: by 2000, when I left, it seemed like everybody was in the game. As Tim Berners-Lee put it, five internet years pass for every year in the real world: by that metric, I spent two subjective decades inside Datacash. My job was to write the server-side software that allowed a Linux box to talk to the British banks’ credit card processing systems (which operate completely differently to the US system). I was under a bit of strain in 1999. Our business was growing at a compound rate of 30% per month, and the code I’d originally hacked out as a proof-of-concept demo was now a mission-critical monster that the company was basing its IPO prospectus on.

(I said I didn’t understand money, didn’t I?)

“Lobsters”, the first story in what became Accelerando, was what I did instead of having a nervous breakdown: I bottled up the angst of acceleration and tried to distil it into a novelette, as a way of explaining to outsiders just what it was like to be inside the internet bubble. Then, in 2000, I began writing a sequel story, because I’d left the characters in “Lobsters” dangling over the abyss of an uncertain future. It took me another four years to finish the process — by far the longest it’s ever taken me to write a novel: three decades of internet bubble-time.

If I understood money, I’d be looking at the current economic situation and licking my metaphorical chops. But you’ll have to find someone else to write you the “Accelerando” of the CDS market.

Posted by Charlie Stross at 10:21 PM | Permalink | Comments (0) | TrackBacks (0)

Science fiction is, more than anything, a literature of ideas. And Charles Stross has more ideas than is probably healthy for one man. How many writers truly grapple with what it is to be human, with or without post-human technology? Accelerando bravely risks alienating you from the characters by propelling them off into multiple iterations far removed from the original meat-space versions. It reminded me of the second half of Wuthering Heights, when the original cast of characters is dead or unrecognizable, and a set of translucent copies play out the same drama. Less satisfying emotionally, but it makes you grasp intuitively the big questions beneath; what is free will? Am I the same person I was before puberty, when I left home, or even this time last year?

Stross often writes about life on the other side of that black hole, the Singularity, a world that is by definition unimaginable. How can we imagine what consciousness, pain or joy might be like after we digitize our brains? Post-singularity writers remind me of Saint Paul trying to explain the transcendental nature of Christianity to a colonized under-class who’d expected the Messiah to literally smite the Romans (and the Egyptians, Persians, and Mesopotamians). There’s a distinctly religious echo to the implication that the ways and thoughts of post-singularity existence are far beyond ours. Who, in their right mind, would even try to write about this? Stross for one. But not only that, he brings on the funny. Stross is a superb comic writer, an absurdist on a par with Terry Pratchett who never slips fully into slapstick: A.I. lobsters, talking telephones, a pitch perfect send-up of communist factions, and my favourite line in perhaps any novel:

“Nobody ever imagined a bunch of Orcs would steal a database table…”

Stross’s stories are always about politics. His characters inhabit fully imagined universes where easily recognizable groups of people grapple with issues and contend for power. They don’t mope around describing the scenery either. Instead of taking half a book to figure out the implications of hopping between universes, Miriam Beckstein is packed and ready for her second trip, and trying to blast a medieval society into the information age by her fourth. Which is great because what’s interesting about fantasy worlds is not ‘how will the lead character get her head around this’ but more ‘but how would it work?’ What’s the plumbing like? Why would Ivy League schooled world-walkers keep their homeland in a feudal state of development? This is more than a fascinating conceit. It makes me understand the House of Saud a little better and reminds me of William Gibson’s famous quote; “The future’s here already. It’s just unevenly distributed.” So is the past.

But while Stross litters his universes with jewels of ideas other writers would lavish novellas on, I wonder if there’s something peculiarly leftist about the revved up short-handing of human progress. Civilizations are tagged pre and post-contact by a technological determinism that drives political, economic and social development on a linear track, albeit at the speed of a geometric progression. Stories abound where future-shocked characters say things like:

“But the UN is a government-“

and are told;

“No it isn’t,” Martin insisted. “It’s a talking shop. Started out as a treaty organization, turned into a bureaucracy, then an escrow agent for various transnational trade and standards agreements. After the Singularity, it was taken over by the Internet Engineering Task Force.* It’s not the government of Earth; it’s just the only remaining relic of Earth’s governments that your people can recognize.”

Iain Banks’ Culture novels are the epitome of the idea that if you magic away resource constraints, politics is about individual identity and the exercise of free will. It’s quite fair to argue that a society where technology has developed so far as to make scarcity unknown, and where digitized humans are impossible to murder, or even, really, to harm; this sort of society would be organized in a radically different way, if it’s organized at all. (Such a society could be described as ‘organized’ only insofar as emergent patterns and associations can be identified, rather than being structurally determined by design or consent.) How useful is it, politically, to speculate about what post-singularity life might look like? The description of elections in Accelerando as the acme of brand-driven, micro-marketed memetics isn’t all that satisfying. It’s politics as we know it, with more processing power and faster cycles. In the post-singularity politics of the Eschaton, people fight for liberty, not resources, because freedom is the only thing there’s a shortage of. But on closer inspection, the main struggle in Stross’s near-future writing is not for survival but for freedom, too.

Halting State is set in a pre-Singularity near-future where the commonest application of AI is as a spam filter that summarises an email as “job offer, vaguely menacing”, 70% likely to be spam, but probably worth a look. Halting State does what near-future SF does best; extrapolates current trends and technologies into a recognizable scenario that critiques the present day. It’s a Britain I certainly recognize.

We still have bendy buses, but the Republic of Scotland is the new Celtic Tiger and uses Euros instead of sterling. The band-aid covered Computer Misuse Act is still going, though with Scottish revisions post-independence in 2014. The lumbering infrastructure of the Regulation of Investigatory Powers Act has shaped society, but the local PC plods are still as hopeless with IT crimes as they are today. The appearance of Euro-spook Mehmet indicates that Turkey may have joined the EU and the present-day power grab of the EU Council of justice and home affairs Ministers paid off. Britain’s slid further down the slippery slope to a surveillance state. Law enforcement have always lived in a different world, and now they inhabit a data-rich version of reality called CopSpace. The Tube is dirtier, and even more screwed up by under-investment and the skewed incentives of public private partnerships. Global warming has made London sweaty from April to November and driven up the cost of flood insurance. Going behind the school bike shed at the age of fifteen with a younger girl puts you on the sex offenders list for life. But at least the cops take Paypal.

It’s all a bit depressing, really. (Especially if you’ve spent a good chunk of your career fighting the expansion of state surveillance and still can’t understand why the UK has the best thinkers, writers and activists on this stuff, but some of the worst policies by far.) But in Stross’s oeuvre, the Singularity will somehow give us a pass on Big Brother’s Brave New World. I’d love to hear how. Of course, the singularity is a qualitative change about more than just faster processing power, and a self-replicating cornucopia machine will put a lot more than the means of production in the hands of the workers. But in a fictional universe where technology drives the politics, how might we get from a pre-Singularity panopticon to a world of free-floating and interchangeable individual, corporate and government identities?

I love that towards the end of Singularity Sky, Rachel channels John Perry Barlow:

“We’ve been trying for years to tell your leaders, in the nicest possible way: information wants to be free. … Then along comes the Festival, which treats censorship as a malfunction and routes communications around it. The Festival won’t take no for an answer because it doesn’t have an opinion on anything; it just is.”

I got involved in Internet policy in the first place because I thought this kind of escapist rhetoric was precisely how the cyber-libertarians were going to sell the farm to big business and repressive governments. The Internet has massive ability to spread knowledge and ideas, and it was designed to route around and rout out single points of failure. Many early adopters, of a decidedly libertarian outlook (what with being young, white, affluent and male), took this to mean the Internet is antithetical to centralized control. The evolutionary ideal of the Internet had it developing antibodies to censorship and undermining authoritarianism in all its forms.

The reality, though, is that the Internet and its associated tools are developing as the ultimate technology of control. Far from being much able to influence developments in the opposite direction, my professional life has just given me a bird’s eye view of the coming train wreck. So, eh… read Charles Stross. (and Neal Stephenson, Greg Bear and Kim Stanley Robinson.) His books help us think through these issues, follow current trends to their logical conclusions, and make it clear that some worst-case scenarios are anything but fiction.

But I’ll leave the last word to Harald Alvestrand, a former chair of the Internet Engineering Task Force who I’m lucky enough to know through my work. A few months ago I asked Harald what he thought of the singularity and when we might reach it. We defined terms, and agreed the singularity might mean the exponential increase in technological progress that takes in computing, nanotech and cognitive science. Harald said the singularity’s already here, it has been for quite a while, and that it’s an exciting time to be alive.

• I especially loved this since so much of my daily work is affected by the efforts of a UN body to take over some Internet numbering/naming functions.

Posted by Maria Farrell at 12:16 AM | Permalink | Comments (0) | TrackBacks (0)