Cyberspace, the Singularity, Belief Circles, oh my!

August 24, 2007 8:01 AM

Vernor Vinge: Mathematician, computer scientist and science fiction visionary worthy of Arthur C Clarke's mantle, Vinge is most famous for popularising the idea of the singularity, where technology advances so quickly that humans cannot participate, but he's also credited with writing one of the first stories about cyberspace, True Names, back in 1981. More recently, he's been exploring how augmented reality and belief circles will change the way we live in his latest novel Rainbows End - which he put online, completely for free.
posted by adrianhon (43 comments total)

Other interesting links worth reading include his essay on What if the Singularity does NOT happen? and Synthetic Serendipity, his short story (essentially an excerpt from Rainbows End) about augmented reality games and what happens to the people who can't keep up with the accelerating pace of technology.
posted by adrianhon at 8:06 AM on August 24, 2007

A Fire Upon the Deep and A Deepness in the Sky are two of my favorite books.

Just wanted to say.
posted by linux at 8:16 AM on August 24, 2007 [2 favorites]

I actually think that Fast Times at Fairmont High, the novela Rainbows End was based on is a lot more focused and generally better.

linux - I hear a follow up to A Deepness in the Sky is on the way BTW. Also if you've not already checked out Iain Banks I would do so, I think anyone who likes those two novels would love his Culture stuff.
posted by Artw at 8:24 AM on August 24, 2007

ditto linux, and I just wanted he's so cool! I had a long chat with him about this stuff at the Accelerating Change conference back in 2005. That reminds me, I thought his next book was going to be about India, that's what he'd said.
posted by infini at 8:24 AM on August 24, 2007

woops that should be "I just wanted to say..."
posted by infini at 8:24 AM on August 24, 2007

I listened to his Long Now talk on the 'non-singularity' possibility, and what struck me was his utter conviction that the singularity would happen. I thought about Clarke's First Law, and mentally composed a prospective 'Vinge's Corollary': "Whenever a distinguished scientist or futurist says something is inevitable, they are very likely wrong."

Of course it's silly to critique SF on its predictive capacity. But I think it's valid to critique it for getting too wrapped up in its own belief in its predictive capacity. Singularitarian writers often seem to me to get quite wrapped up in whether or not it's likely to happen, when its nature (if it's indeed a singularity, if that's an apt metaphor) is such that you just won't be able to know.

The Singularity as an idea is really powerful and opens the door to a lot of thought experiments that are very provocative. It allows us to think about a post-human future in new ways. But in practice, it ends up often being more like a religion -- and, paradoxically, often not really employing much imagination. Do I really think a post-singular being is going to think and act like humans? Why would I think that? Why would they care?

I found Sterling's take [slideshow] on the whole thing kind of refreshing. (Podcast version, also Long Now.)
posted by lodurr at 8:30 AM on August 24, 2007 [1 favorite]

lodurr: I agree that a lot of the Singularitarian writers get carried away, but what I've found about Vinge's stories are that they're more about the effects advanced tech will have on humans; the singularity doesn't really factor much into his most recent books. In fact, from a predictive point of view, Rainbows End is much more interesting because it's a near-future book where all the tech discussed is perfectly possible right now. Silent messaging, for one, is doable and will likely have an even bigger impact on our behaviour than mobile phones and SMS combined.
posted by adrianhon at 8:42 AM on August 24, 2007

Of course it's silly to critique SF on its predictive capacity.

Arthur C. Clarke on his predictive abilities. "Of all the things that I guessed would be around in 2001, I thought Pan-Am and The Bell System were the most certain."

(Quote is by memory, but the substance is correct)
posted by eriko at 8:45 AM on August 24, 2007

One of my treasured books is a first edition paperback of "True Names", found at a Half-Price Books in Austin five years ago after checking for it once a week for YEARS.
posted by mrbill at 8:51 AM on August 24, 2007

I also want to say that A Fire Upon the Deep and A Deepness in the Sky are two of my favorite books, that I have ready many, many times. Rainbows End didn't wow me the way it wowed others on the first reading, but I got a lot more out of it on the second reading. Long live Vernor Vinge!
posted by Divine_Wino at 8:53 AM on August 24, 2007

True Names, baby, True Names...
posted by Samizdata at 9:04 AM on August 24, 2007

I was supposed to go out to a movie the day I finished "Marooned in Realtime". I finished it in the theater by the lights of the opening credits.
posted by Kid Charlemagne at 9:06 AM on August 24, 2007

It's worth noting that the reason he has made it available is because it is nominated for the Best Novel Hugo, which will be announced next week. The Hugos have also got a shiny new website.
posted by ninebelow at 9:07 AM on August 24, 2007

Vernor Vinge is one of the few authors great enough to share a genre with PKD.
posted by smackwich at 9:31 AM on August 24, 2007

True Names looks pretty cool, but the "short stack" is pretty tall right now and a lot of stuff would be ahead of it. The title evokes mystical modes of interaction that have fascinated me ever since I first encountered them. I don't know if this is what he's interested in with that book, but I'm very interested in the ways that technology enables us to be less and less rational.

One possibility of ubicomp that's fascinating to me is that as we diminish the technical arcana, we enhance the mystical arcana -- that is, as we make the interfaces more metaphorically transparent, we are actually increasing our remove from the machine.

That has ramifications in turn for how we deal with facts in the world. If we're accustomed to dealing with abstractions of reality -- we gesture to "pick up" our "mail" and "open" it, as a trivial example -- then the real reality becomes removed. This gets serious when we have a problem really understanding, say, what it means to be hungry or thirsty or cold, when those conditions are applied to someone living in a place we're not logistically obligated to care about.
posted by lodurr at 9:37 AM on August 24, 2007 [1 favorite]

BTW: At least one other person around here has something up for consideration, too. (Others?)
posted by lodurr at 9:41 AM on August 24, 2007

Oh man, Marooned in Realtime. I never even read the book, I just kept skipping around to find the parts where the person was marooned, then to the end to find out what happened. What a nightmare scenario.
posted by DU at 9:51 AM on August 24, 2007

So let's look at the real-world technological end of the "localizers" in the augmented reality link, just because I actually built something applicable to that.

Basically, you can do a lot of cool stuff if you have small devices that can know exactly where they are, especially if they can perform this operation many times a second. (E.g. you could do easy and versatile motion capture for video by sticking a bunch of these on someone if they would update 30 times a second, and that's one of the lesser applications.) For real world use this will need to be robust and versatile enough to perform in many environments - so GPS is right out, as is anything that will definitely fail when your toy robot drives under the bed or around the corner. You also want it to be accurate - each level of accuracy you attain will enable more and more applications. Accuracy of a millimeter is better than a centimeter is better than a meter. You want it to work in three dimensions and you don't want it to include absurd amounts of infrastructure (so specially wired floor tiles, which has been done, are right out.)

As far as I know, this is not a practically solved problem for accuracies less than a foot order of magnitude, even in the lab. The approaches I know of are either based on acoustics (you can use audible or ultrasonic) or ultrawideband radio transmissions. I have a hunch the real success is likely to come from the UWB technology, which I've read less about.

So, you have signal emitters and signal receivers. Which is on your fixed-position beacons and which is on your mobile devices is pretty irrelevant to the math, and of course you could have devices with both and have the devices themselves do a lot of the beaconing, though chances are you'd probably still want some fixed reference beacons on/in your ceiling to avoid building up too much error. A signal or signals are emitted and received. Either the mobile device sends out a signal which is received at several fixed beacons or several beacons each send out a signal, which is received at the one mobile device. Then, using the received signals, you try to figure out geometric properties that, combined, let you figure out the location of the mobile device. For example, if you can figure out how long the signal took to arrive and you know the propagation speed of the signal, you have a distance measurement between a fixed station and the mobile device. Get enough of those and you can trilaterate. A common approach is to use "time difference of arrival" (TDOA) measurements - the signal reached station A x milliseconds (or nano/picoseconds, for EM) before it reached station B, which after a lot of math that some poor bastard did once, turns out to define a hyperboloid (approximate as the surface of a cone) on which the mobile device must be located. Intersect enough hyperboloids, and you get the location. There's enough other methods - something called beamforming, for example, could be applied to this problem.

You can make things easier with assumptions and accomodations - if your fixed beacons and mobile devices have synchronized clocks and a wireless data connection (possibly the same as the localization signal), you can know that the signal was emitted at time t as well as knowing that it was received at time t+x, which allows easy distance measurements. When you don't know when the signal was emitted, you might have to resort to TDOA.

In my particular project, the mobile devices were actually arbitrary sound sources, mostly people speaking - so we were working with no technological accomodations on the part of the "signal sources." The theoretical end-game of our particular project would have been this: Consider a party with tons of people talking at once. Hang enough microphones on the ceiling - theoretically, one per person, practically, more - and you could generate a recording of each person's voice, as if they were wearing a lapel microphone - even if every ceiling mic heard a mixture of everyone's voice, and locate each person. That first part is called the "cocktail party problem" and the solution my teammate had for it was real magic (involving looking at the signals with a lot of probability stuff and assumptions and somehow it works) even compared to the localization stuff.

The biggest practical problems for the acoustic localization were reverb, which just destroys TDOA accuracy and can't be eliminated since sound bounces off anything, and range - we had it working in a small room sized area within a larger room, i.e. we weren't too close to walls or too far away from the mics. I think the range was more a problem for the separation, a prerequisite for the localization, but we didn't look into that. But within this small room area, we were locating with I think roughly 15 cm errors with no accomodations on the part of the signal sources - in a practical localizer system your signal sources would be small devices and would be able to make accomodations. We also weren't working in real-time. We didn't try, just post-processed recordings in MATLAB, but it was up in the air whether or not a faster implementation would have handled real-time, although it was close enough that a few years' progress should certainly allow it.

But basically, we were college seniors working with 16 boring microphones and a boring, slightly out-of-date computer, and we were pulling this off, so this technology is coming.
posted by TheOnlyCoolTim at 10:12 AM on August 24, 2007 [2 favorites]

Also, the coolest localization project to look at in terms of videos and stuff you download is MIT Cricket, though they seem to have sort of died down since someone got his doctorate.
posted by TheOnlyCoolTim at 10:17 AM on August 24, 2007

From the Wikipedia link: ...drawn in a complex plot involving a traitorous intelligence officer, an intellect of frightening (and possibly superhuman) competence hiding behind an avatar of an anthropomorphic rabbit...

Algernon?
posted by localroger at 10:25 AM on August 24, 2007

Of all the things that I guessed would be around in 2001, I thought Pan-Am and The Bell System were the most certain.

Well, I think one of those predictions is gradually coming around.
posted by zek at 10:29 AM on August 24, 2007

Overheard at the opening of Tom Cruise's new movie Algernon: 2025: "The thing I hated most is that they changed the mouse into a rabbit."
posted by localroger at 10:29 AM on August 24, 2007

A Fire Upon the Deep was a good read, but I really could've done without the sentient dog sex scene at the end that forever lodged this book firmly in the Furry canon.
posted by solipsophistocracy at 10:34 AM on August 24, 2007

Singularity = Rapture for Nerds. (Thanks, Ken!)
posted by Justinian at 11:14 AM on August 24, 2007

The only thing more certain than the ironclad faith of Singularitarians in their Computronium Rapture is the astonishing regularity of Vinge repostings. Having said that, I recently re-read all of Vinge, even the naff shorts in the collected stories. Someone I know worked in publishing when DitS was being delivered - I gathered from them that part of the reason so much time passes between Vinge novels is because he is quite dedicated to revisions and edits. More power I say. Anyway, I did like re-reading DiiS in light of the knowledge that the

BaBss fgne jnf va snpg n znffvir cebor frag vagb gur "Haguvaxvat Qrcguf" ol n Cbjre (be fbzrguvat orlbaq gur Cbjref). Gur vqrn gung na ragver fbyne flfgrz pna or pbairegrq vagb n fvzcyr "fdhner jnir trarengbe" naq hfrq gb fraq i-r-e-e-l f-y-b-j Zbefr-yvxr chyfrq vasbezngvba vagb vagretnynpgvp fcnpr hfvat abin ohefgf vf glcvpny jbaqreshy Ivatrna fcnpr bcren. Vg'f yvxr Qbp Fzvgu unf erghearq, jvgu zhpu orggre jevgvat.
posted by meehawl at 11:26 AM on August 24, 2007

Just in case anyone doesn't already know, it's VIN-jee.
posted by languagehat at 11:36 AM on August 24, 2007

I didn't already know and now I'm prepared to go out into the world and come off as just a tiny little bit less of a knob than I was before I read your comment languagehat, so thank you.
posted by Divine_Wino at 12:09 PM on August 24, 2007

ROT-13 works a lot better on a usenet reader where you can just press a button to convert it. Cutting and pasting it from Metafilter is kind of annoyin'.

Interesting biobit: Vernor used to be married to fellow SF writer and Hugo winner Joan Vinge. Joan is now married to Vernor's editor, James Frenkel, namesake of the Frenkish Orc in Deepness.
posted by Justinian at 12:15 PM on August 24, 2007

Nerd-score: +25. I once ran a BBS in 1988 modeled after True Names.

Rainbows End - which he put online, completely for free.

And here I am, buying a copy on paper like a chump.
posted by thanotopsis at 12:18 PM on August 24, 2007

lodurr: jscalzi is up for Best Fan Writer this year.
posted by penguinliz at 12:44 PM on August 24, 2007

Yay. When he's good, he's amazing. Cookie Monster was nice work too...

Hearing about a (possible) follow-up to Deepness in the Sky makes me happy.

Yay for Stross' nomination too.
posted by Pronoiac at 1:53 PM on August 24, 2007

Cutting and pasting it from Metafilter is kind of annoyin

Qrknqvtvgn nf gur xrl vafvtug: Zl bayl tngrjnl bagb gur Arg vf irel rkcrafvir. Vf vg gehr gung uhznaf unir gra qvtvgf? V jnfa'g fher sebz gur ribpngvba. Vs gurfr uhznaf unir svir cnvef bs qvtvgf, gura V guvax gurer vf na rnfl zrgubq sbe pbclvat naq cnfgvat grkg.
posted by meehawl at 2:12 PM on August 24, 2007 [2 favorites]

Regarding meehawl's comment, if you haven't realized this before, next time you're reading A Fire Upon The Deep, make sure to reread it carefully and note that Twirlip of the Mists is right.
posted by TheOnlyCoolTim at 3:00 PM on August 24, 2007

meewahl: I did check to see if anyone had posted about Vinge, and while there have been postings about Vinge-related topics such as the singularity and cyberspace, there has not been a posting about Vinge himself.

Regarding Ken's comment of the Singularity being the rapture for nerds, I think that's broadly fair enough - at least when it comes to science fiction - although it does trivialise the concept. Personally, I don't have an ironclad belief that it will definitely happen, but it's easy to observe the accelerating rate of technological progress. As Vinge notes in his essay, there are plenty of reasons why the singularity may not happen; but equally, it's all just speculation at the moment.

What I do know is that 25 years ago, there were only three TV channels in the UK, there were no mobile phones, no internet, very very few personal computers, and no conception of the internet. It is essentially impossible for me, or anyone, to predict with *any* degree of confidence what sort of technology we will have in 25 years time - and that's an exciting and scary thought.
posted by adrianhon at 3:32 PM on August 24, 2007

Side-note: I cycled up the "Snake Path" by the UCSD library from RE two days ago, and while there read the relevant passages on my 1.5 Mbps HTC PDA phone by pulling down the etext from Vinge's vrinimi website and doing a quick grep. Then I pulled down the Fire Upon the Deep etext and re-read the bits that mentioned the "vrinimi" galactic ISP. That is a kind of progress, but it doesn't really make me feel singularly superhuman, but more vaguely super-nerdy.
posted by meehawl at 3:55 PM on August 24, 2007

I'm kind of sad that's how he pronounces his name, 'hat. I always assumed it was a nom de plume/portmanteau from verge and whinge ... which seemed so apt.
posted by rob511 at 3:56 PM on August 24, 2007

Right, Rapture of the Nerds.
posted by Freaky at 5:02 PM on August 24, 2007

Tough Guide to the Rapture of the Nerds: What you need to understand is that after the SIngularity things will be cool. And there'll be jam for tea every day.

(meehawl's comment reminded me of a /. username - anyone else?)
posted by Pronoiac at 5:32 PM on August 24, 2007

Re AcceleratingFuture.Com, methinks the nerd doth protest too much. "*We're* rational, those guys are nuts", "They do ritual and have holy writings, we don't", "They are passive wrt their Gods, we're not", etc etc. All these unprovably false assertions incorporate a veracity that depends only your cultural framing. MacLeod's nuanced criticism and engagement of Singularity as a trope within SF is quite more complex than the "Rapture" soundbite, and has been rather formally critiqued. Impassioned Marxists, Trotskyists and Libertarians are always eager to claim progressive, historical rationality, as they do in MacLeod's stories... See also, passim.
posted by meehawl at 5:38 PM on August 24, 2007

Meehawl: I'm not sure that I see what you mean. The AcceleratingFuture post is not perfect, but I think it addresses a lot of the criticisms people have of 'singularitarians' - and I see we've moved on from the whole 'it's just for nerds' to 'it just can't happen'.

I quite like MacLeod's stuff and I've read most of it, but I can't say that I thought it was nuanced criticism of the singularity; there was little discussion of the possible technology involved (I mean, clockwork computers?), and more about political systems, as you say. I looked back at your other comments and it seems you discount the singularity based on the fact that cheap access to energy will soon disappear.

I don't buy it. Energy is a *serious* problem but we are already seeing the European developed nations reducing their power consumption and switching over to renewable energy. Of course, the real issue is the developing world, but if you think that technological development will suddenly grind to a halt because we don't have any oil left, I think you're absolutely wrong. I'm actually surprised at how quickly the switch from fossil fuels is taking place now.

I also don't agree that acceleration is confined to western, developed nations. Mobile phone usage - a magical device only 20 years ago - is blossoming across all nations, and you only have to look at countries like China and India to see that the west is no longer alone.

In any case, I think we do agree in some points. There could be a big war. There could be a huge pandemic (although our ability to avoid such things is increasing). No-one doubts that, and in that case, sure - no singularity.
posted by adrianhon at 2:03 AM on August 25, 2007

and has been rather formally critiqued.

I don't remember any of those essays engaging primarily with the Singularity.
posted by ninebelow at 3:37 AM on August 25, 2007

it addresses a lot of the criticisms people have of 'singularitarians'

Actually, it really doesn't. You can *say* 'we're not like those guys' until you're blue in the face, but from an outside perspective, well, you are. "Sacred texts" vs "The Singularity Is Near"... you decide.

switching over to renewable energy

The problem with renewable energy is that, while it's renewable, by some estimates we are currently consuming more than 400 years of the Earth's Net Primary Productivity every year, thanks to the wonderful capacity of fossil fuels to compressively bank millennia of solar input within a single relatively narrow stratum. So it renewas at a rather slow rate.

Now, fast breeder reactors give you a little win here, and tinkering with the genetics plants and maybe improving silicon a little gets you some juice. As does geothermal, and looting methane hydrates. But there are constraints here - the breeders are a relatively short-term solution, the plants are already pretty well optimised little nano assemlbers, and creating lots of photovoltaic silicon is massively consumptive of rare water resources.

Basically, you get slight linear improvements, whereas your demand is exponential. You want fusion? Looks like you're going to need He3, which is hard to come by in mass quantities - you want serious amounts you're going have to set up extraction plants in incredibly toxic and dangerous extraterrestrial environments and pull it up out of steep gravity wells.

technological development will suddenly grind to a halt

No, it won't, mainly because oil and gas for fertiliser won't drop off a cliff but simply become less available and less economically sustaining. You'll see more of what you're seeing now: differential inflation in food sources and animal derivatives, and a re-alignment of politics to re-favour the commodity nations.

However, as energy sources become increasingly marginal, I suspect a process analogous to that of Europe in the 13th and 14th centuries or China in the 18th and 19th centuries will emerge. After several centuries of impressive growth, those societies reached a peak of population and development, having deployed massive water and wind energy extraction technologies. They embarked on huge, decades-long engineering and colonisation projects to sustain their economies and engage their surplus populations. However, their situations became increasingly unstable, and in Europe's case was destroyed by climate and then a pandemic, and in China's case climate followed by a Christian-inspired civil war of dramatic intensity.Jones's The European Miracle has a lot to say on the divergence between Europe and China in the 18th and 19th centuries, and much of it has to do with Europe's successful appropriation of vast new agricultural and fossil fuel reserves outside its territories. What's refreshing is that it includes a lot of well-researched economic figures, unlike the general hand waving of many successor books.

Hopefully, our case will resemble more a simple, long-term economically deflationary period out of which will emerge a lower-intensity, rather solipsistic culture using augmented reality and telepresence to substitute for a lack of cheap energy to move bodies around en masse (which is what we seem to be currently obsessed with).

I don't remember any of those essays engaging primarily with the Singularity.

Well, I think that Scottish writers in general tend to be a little more critical of faith in unvarnished, unidirectional progress. The Scottish Watt basically invented the modern era by devising the first truly efficient steam engine that was an improvement over the medieval renewable energy-driven industrial machines. Scotland was way ahead of most of the world for a while and then, well, it wasn't. Progress is not inevitable.

MacLeod's fiction as a genre tends to engage with the Singularity because of his perception of it as a classic Hegelian Historicist narrative. That's why his Marxist protagonists always relate to well to it. It's dialectical technologism writ large, a kind of Nerdish Idealism prioritising a rather materialistic idea of transcendence. As a motivating factor, Eros plays a large part wherein many people have sublimated into the Singularity a great reluctance to face the fact of their impending death. Hence also Kurzweil's brisk sales of nutritional supplements. Taken as a whole, the Singularity Circus also makes easy TV candy.

Singularitarians love showing off all these exponential graphs, which are based on dodgy data, rarely have error bars, and to me seem to represent a measurement of our economic consumption more than anything else.

The idea that reliance on little blinking gadgets will enable us to transcend our natures and environmental limits is endearing. However, I have read many of the scifi and hobbyist mags of the 1920s and 1930s. It's remarkable how many of them fervently believed that radio was going to usher in a new world of unbridled prosperity. Their faith in the magic little box that could deliver improving thoughts from one person to many all over the world is resolute and there are rarely dystopias. What did we get? Goebbels.

Finally, it's worth ending with a quote from Jung on Hegel's progressive optimism:

It is reminiscent of the megalomaniac language of schizophrenics, who use terrific, spellbinding words to reduce the transcendent to subjective form, to give banalities the charm of novelty, or pass off commonplaces as searching wisdom. So bombastic a terminology is a symptom of weakness, ineptitude, and lack of substance.

posted by meehawl at 10:01 AM on August 25, 2007 [2 favorites]

adrianhon: It is essentially impossible for me, or anyone, to predict with *any* degree of confidence what sort of technology we will have in 25 years time

And yet, people today live in ways that are more or less scalable from the way they lived 25 or 50 years ago. Some things remain consistent, even through the vast technological changes of the past 200 years. There are some essential truths of human nature that I dont think most singularity fiction ("SF"?!) deals with all that well. In a nutshell, in my opinion most singularity fiction has humans (and "uplifted" beings) behaving too much like they do now, or not enough.

There's a great line from Neuromancer (the Finn, speaking) to that point -- my copy is lost in the morass of my stepson's room, so I can't cite it, but it goes something like this: Don't trust AIs, because they're not human and you can't possibly understand their motives. That is, even when they seem to behave like humans, it's an illusion, perpetrated for the sole purpose of making people think AIs can be understood. I get that that's part fo the point of the Singularity, but I don't think most people who are enamored of it get it, even as they claim to.

That said, there are of course a number of basic problems in writing about a being who's truly, deeply different from your readers. Not least among them is getting people to read it. And I also believe that at some level, science fiction is basically a subset of fantasy (in the sense in which Ursula le Guin uses the term 'fantasy') -- at its best, it's a thought experiment that might trade as much in metaphor as in speculation. The best SF leaves you thinking; hell, some stuff I really dislike in one way or another leaves me thinking a lot. For example, I could sit here and shoot holes in Cory Doctorow's singularity-focused stories, for hours, but at the end of the exercise all that would have really been demonstrated is that Cory provoked me to think about all those things. Put another way, I may have real issues with the details of the picture he paints, but dammit, I looked at it and thought about it. That's worth something.

But I realize that I digress, again.

Singularitarianism seems to me to be something that's relevant only to a relatively small part of the human race -- unless, of course, the singularitarians become so powerful that they end up determining the fate of everyone else in the world. I don't find that likely; there's just too many people in the world and there's too much social inertia. I haven't read MacLeod's [Ken MacLeod?] critique, so maybe that's what he's saying -- that sounds like a "rapture of the nerds" to me.
posted by lodurr at 1:58 PM on August 25, 2007