Twenty years of Taklamakan

Bruce Sterling’s short story was published in 1999.

I figure he must have written it at least twenty years ago now.

I still think of it multiple times every year.

Severe annual resonance increases.

NAFTA, Sphere, and Europe: the trilateral superpowers jostled about with the uneasy regularity of sunspots, periodically brewing storms in the proxy regimes of the South. During his fifty-plus years, Pete had seen the Asian Cooperation Sphere change its public image repeatedly, in a weird political rhythm. Exotic vacation spot on Tuesdays and Thursdays. Baffling alien threat on Mondays and Wednesdays. Major trading partner each day and every day, including weekends and holidays.

At the current political moment, the Asian Cooperation Sphere was deep into its Inscrutable Menace mode, logging lots of grim media coverage as NAFTA’s chief economic adversary.

As far as Pete could figure it, this basically meant that a big crowd of goofy North American economists were trying to act really macho.

Their major complaint was that the Sphere was selling NAFTA too many neat, cheap, well-made consumer goods. That was an extremely silly thing to get killed about. But people perished horribly for much stranger reasons than that.

At sunset, Pete and Katrinko discovered the giant warning signs. They were titanic vertical plinths, all epoxy and clinker, much harder than granite. They were four stories tall, carefully rooted in bedrock, and painstakingly chiseled with menacing horned symbols and elaborate textual warnings in at least fifty different languages.

English was language number three.

“Radiation waste,” Pete concluded, deftly reading the text through his spex, from two kilometers away.

“This is a radiation waste dump. Plus, a nuclear test site. Old Red Chinese hydrogen bombs, way out in the Taklamakan desert.” He paused thoughtfully. “You gotta hand it to ’em. They sure picked the right spot for the job.”

“No way!” Katrinko protested. “Giant stone warning signs, telling people not to trespass in this area? That’s got to be a con-job.”

“Well, it would sure account for them using robots, and then destroying
all the roads.”

“No, man. It’s like—you wanna hide something big nowadays. You don’t put a safe inside the wall any more, because hey, everybody’s got magnetometers and sonic imaging and heat detection. So you hide your best stuff in the garbage.”

 

Station Identification: Rule 110

“For the ‘One Hundred Billion Sparks’ album project I want to tell a story of our one hundred billion sparking neurones, and the magic which they create: our minds. Early in the story I aimed for the “nuts and bolts” of the processes involved, but not in the sense of showing a neuroscience lecture, I want to find the artistry and beauty of the natural processes involved.

Those are what make the richest visuals for my videos and live shows. Following this reasoning, one idea which came along, was to visualise a “Turing-complete” machine, which is a computer that is capable of performing any computation. This means the design of the computer is versatile enough to allow for any logical operation, within the constraints of the sorts of logical operations our usual computers can do. David Deutsch, amongst others, makes a convincing argument that human brains must also be universal computers in this sense, in his interesting new book ‘The Beginning of Infinity’. So I have some rough grounds at least, for making this link between brains and computers for the purpose of trying to get some hint of the visual essence of thought.

The interesting aesthetic link comes in via the work of Stephen Wolfram, from his 2002 book, ‘A New Kind of Science’, where he shows that simple “cellular automata” models, growing blocks of binary colour following simple rules, can create rich behaviours in their growth patterns, and even yield a system capable of Turing-completeness. Following a systematic exploration of the simplest possible rules governing cell duplication, Rule 110 is the first rule which displays Turing-completeness and is the simplest visual system that I know of which embodies this attribute.

The really interesting thing is that Rule 110 also displays a very particular visual aesthetic, that of a combination of order and chaos, never totally predictable or totally random. For me, that potential artistic/aesthetic link to universal thought is pretty amazing, and it’s also an aesthetic/property which appears in many other important places in nature (for example https://maxcooper.net/the-nature-of-nature), as well as being one of the main principles of my approach to music, where a healthy dose of disorder is always important.

After settling on this visual form for the project, I needed to create a piece of music which suited the retro blocky nature, which is something akin to Tetris. My immediate thought was big gated reverb snares and powerful classic synths. It had to be bold and clean in one the large scale, but also full of generative unpredictability.

It all fit nicely with what I like to do anyway, and just pushed me in a slightly more poppy direction than anything else on the album. The initial focused time was spent finding the killer chord sequence and bold patch, then setting up a generative seething chaos of synthesis with plenty of random waveforms and modulations, then a long time on the arrangement detailing with more than 100 layers of sounds. I finally added a vocal from Wilderthorn, which I chopped into destruction, just there to add a little hint of humanity in amongst the computation.

The final step in the process was to chat to the great visual artist, Raven Kwok, about the ideas and what I would like from the video. I was really happy when Raven showed me that he wasn’t just going to make an artistic interpretation of Rule 110, but had actually built his own version of the real system!

So the video shows an authentic pattern-generation of Rule 110, where we can see moments of repetition and pattern, but never in perpetuity, it always returns to disorder. The colours and 3-dimensional explorations are Raven’s extension of the basic system.

I still find it counter-intuitive that a simple deterministic system like this can yield undecidability in the content of its output, and I find it inspiring that this property relates to universal computation. It seems to me, at least, like the finest artistry.”

User #821 signing off

About a month ago, I woke up and decided I wasn’t getting any value out of using Twitter.

In fact I was getting negative value – it just made me angry, upset and stressed out.

So I stopped using it and deleted my account.

I was user number 821… I joined back in 2006 and it was a lot of fun for a lot of the twelve years I used it, but then it really, really wasn’t. So far, I don’t miss it that much.

Screen Shot 2018-08-09 at 9.37.32 AMScreen Shot 2018-08-09 at 2.07.08 PMScreen Shot 2018-08-09 at 1.59.07 PM

Speaking my brains about future brains this year

Got some fun speaking gigs lined up, mainly going to be talking (somewhat obliquely) about my work at Google AI over the last few years and why we need to make centaurs not butlers.

June

August

November

Then I’ll probably shut up again for a few years.

Station Identification: Cake tastes better on the moon.

“Did you know cake tastes better on the moon? If you went to Earth and had cake you’d be so disappointed. It’d be flat and heavy and solid. It’s to do with pore size and crumb structure, and crumb structure is so much better on the moon. Every cake you make is three kinds of science: chemistry, physics and architecture. The physics is about heat, gas expansion and gravity. Your raising agents push up against gravity. The less gravity, the higher it raises. You might think, so, if lower gravity makes for better crumb structure, wouldn’t the perfect cake be one you made in zero gee? Actually, no. It would expand in all directions and you’d end up with a big ball of fizzing cake mix. When you came to bake it, it would be very difficult to get heat to the centre of the cake. You would end up with a soggy heart.”

Look for the Lungfish

Charlie Stross on why he mainly no longer reads science fiction books.

The exercise of substituting “SF” for “Design” or “Speculative Design” is left to the reader.

Similar to the sad baggage surrounding space battles and asteroid belts, we carry real world baggage with us into SF. It happens whenever we fail to question our assumptions. Next time you read a a work of SF ask yourself whether the protagonists have a healthy work/life balance. No, really: what is this thing called a job, and what is it doing in my post-scarcity interplanetary future? Why is this side-effect of carbon energy economics clogging up my post-climate-change world? Where does the concept of a paid occupation whereby individuals auction some portion of their lifespan to third parties as labour in return for money come from historically? What is the social structure of a posthuman lifespan? What are the medical and demographic constraints upon what we do at different ages if our average life expectancy is 200? Why is gender? Where is the world of childhood?

Some of these things may feel like constants, but they’re really not. Humans are social organisms, our technologies are part of our cultures, and the way we live is largely determined by this stuff. Alienated labour as we know it today, distinct from identity, didn’t exist in its current form before the industrial revolution. Look back two centuries, to before the germ theory of disease brought vaccination and medical hygeine: about 50% of children died before reaching maturity and up to 10% of pregnancies ended in maternal death—childbearing killed a significant minority of women and consumed huge amounts of labour, just to maintain a stable population, at gigantic and unpleasant (to them) social cost. Energy economics depended on static power sources (windmills and water wheels: sails on boats), or on muscle power. To an English writer of the 18th century, these must have looked like inevitable constraints on the shape of any conceivable future—but they weren’t.

Similarly, if I was to choose a candidate for the great clomping foot of nerdism afflicting fiction today, I’d pick late-period capitalism, the piss-polluted sea we fish are doomed to swim in. It seems inevitable but it’s a relatively recent development in historic terms, it’s clearly not sustainable in the long term. However, trying to visualize a world without it is surprisingly difficult. Take a random grab-bag of concepts and try to imagine the following without capitalism: “advertising”, “trophy wife”, “health insurance”, “jaywalking”, “passport”, “police”, “teen-ager”, “television”.

SF should—in my view—be draining the ocean and trying to see at a glance which of the gasping, flopping creatures on the sea bed might be lungfish. But too much SF shrugs at the state of our sea water and settles for draining the local aquarium (or even just the bathtub) instead, or settles for gazing into the depths of a brightly coloured computer-generated fishtank screensaver. If you’re writing a story that posits giant all-embracing interstellar space corporations, or a space mafia, or space battleships, never mind universalizing contemporary norms of gender, race, and power hierarchies, let alone fashions in clothing as social class signifiers, or religions … then you need to think long and hard about whether you’ve mistaken your fishtank for the ocean.

And I’m sick and tired of watching the goldfish.

Happy 2028!

Azeem’s excellent Exponential View newsletter this week published some predictions for 2018, which you should go take a look at.

A couple of weeks ago in a late night fugue of amateur futurism I sent an email to a few friends, looking a little bit further out and laden with bias / fiction / wishful thinking. Anyway – putting it here to re-examine in ten years time…

1) Hard Brexit happens to the UK by default, rather than planning.

2) UK becomes virtual client state of EU anyway, any business wanting to make money in goods or services (apart from hardliners like Witheredspoons and Dyson) has to ditch ideology and comply with EU regs, without any of the benefits to citizenry. Well done everyone.

3) USA becomes increasingly insular, either as result of second Trump term or more likely to concentrate on recovering from the first… GAFA gets bashed a lot over next 5yrs and plateaus, net-non-neutrality locks in incumbent value chains but favours big carriers/media not big tech.

4) GDPR and other citizen-centred regulation in EU (plus net-non-neutrality) push value models to ‘federated-edge’, Europe becomes IoT2.0 leader, esp around energy, manufacturing, logistics, auto/vehicular. Neural architectures start to dominate, blockchain-like federation of devices lowers reliance on centralised models of computing (and business)

5) European (esp French, German) startups focus on EU and Africa, to an extent Asia Minor and South Asia as markets. Edge-computing hybrids leapfrog solely cloud-centric business models in hipness if not value (yet).

5.1) EU open banking and finance laws + mobile money innovation around Africa make very attractive markets for edge-focussed fintech startups. The Swiss go all-in for this…

6) one of the Gulf states cashes out on oil, goes all-in on becoming a silicon manufacturer/designer/licenser, with focus on neural architectures, Indian and Chinese manufacturers license, buy or copy.

6.1) Tim Cook moves Apple corp HQ to Dubai. Significant automation in Apple supply chain + IP risks allow edge manufacturing of most of their premium hardware. Other members of GAFA pay attention as the fruit company leaves the plateau of the last few years behind.

7) Chinese firm announces it has an AGI running on a supercomputer running on UAE-designed chips.

8) Bezos announces Amazon to HQ in Zurich (and an orbital solar-powered blockchain datacore at LaGrange Point L1)

9) It’s 2027 and 60% of tech ‘unicorns’ are HQ’d in EU, UAE, India or Singapore.

10) UK joins special economic area of EU in order to adopt data laws formally. Press starts to refer to this as Brentrance…

Lightpainting Centaurs

If you (or anyone) still read this you’re probably aware I’ve been banging on about Centaurs for a little while.

I started idly sketching something that could become a shorthand for a ‘centaur’ actor in a system. The kind of visual shorthand that you might often use on whiteboards or in sketches of flows in designing interactive systems.

For example… back in 2006 I sketched this…

My first centaur symbol sketches… was trying to make it something quick and fluid but kept getting hung up on the tail…

 

I then progressed to subjecting colleagues (thanks Tim) to impromptu lifesize whiteboard centaur sketches…

IMG_20170830_114031.jpg

But then I remembered Picasso’s 1949 light paintings of centaurs  which inspired me to do some quick long-exposure experiments.

 

IMG_1069.jpg

Again, long-suffering colleagues were pressed into service (after buying them some beers…)

And I think that the constraint of having to paint the centaur body in a few seconds of long exposure got me to a more fluid, fluent expression

IMG_1077.jpgIMG_1079.jpgIMG_1080.jpgIMG_1081.jpgIMG_1082.jpgIMG_1083.jpg

But then I think in the end it was Nuno who nailed the tail on this old designer…

IMG_1084.jpg

More centaurs soon, no doubt.