In 2006 Netflix, at the time an online DVD rental company, ran a competition to improve their recommendation system. Their existing recommendation system, called Cinematch, made predictions about how many stars a user would give a movie, based on how they had the rated movies they had watched before. If Cinematch thought a user would give a film a high star rating, and the user had not already seen it, then Netflix would promote that movie to the user as their next rental choice.

They wanted to improve on Cinematch, so in 2006 they offered a prize of $1,000,000 to any team that could improve the accuracy of predictions by 10%. Cinematch was not a sophisticated system, and teams in the competition quickly worked out how to get a 7% improvement [1].

Getting the full 10% improvement was significantly more difficult. Eventually the prize was claimed in 2009 by a team called BellKor’s Pragmatic Chaos, a collaboration by several competitors who combined their techniques to get over the 10% line.

During the three years the competition ran Netflix had moved on, the system was never implemented. A sequel to the competition was cancelled amid privacy concerns – Netflix had claimed that the data it released about star ratings as part of the first competition was anonymised and could not be tied back to any individual user, but two researches had from Texas proved this claim to be false [2]. However, the company still felt Big Data was going to be the key to their future.

The process of recommending a movie for users can be broken down into two steps. The first is to discover what kind of movies the user likes, the second is to find a movie that fits this category which the user hasn’t seen. What happens if your system says users want to watch, say, a Washington-based big name political drama and they have watched all such existing material?

One answer is to make more, which is what Netflix did in producing the TV series House of Cards a political drama featuring Kevin Spacey and directed by David Finch . While the precise impact of Big Data on the production process of House of Cards is unclear, Netflix certainly gained much press attention for their use of extremely detailed data about movie consumption when deciding what kind of show to make [3]. As a result the media have taken House Of Cards to presage a statistical era of cultural production where Big Data is king, with headlines such as ‘How Netflix is Turning Viewers Into Puppets’ [4], ‘“House of Cards” and Our Future of Algorithmic Programming”’ [5], ‘The Secret Sauce Behind Netflix’s Hit, “House Of Cards”: Big Data’ [6]

In this paper, three problems with the use of Big Data when making cultural artefacts will be discussed:

1) Big data approaches struggle when failure is expensive

2) Cultural output may become boringly repetitive if statistical methods are applied

3) Cultural output is reflexive, and relies on a shifting context that might be hard to capture by analysing existing behaviour

As a result, big data is most likely to be used as a rough guide, rather than to completely specify, cultural artefacts – unlike the way it has been used in other industries – since most creative endeavours already use informal historical data to forecast success anyway. For these reasons, I suggest that In contrast to other arenas, in culture big data will only represent an incremental change in approach.

The need for human intervention

It’s hard to discover exactly how instrumental the use of ‘big data’ was in the the formulation of House of Cards, how much of the programme’s success is due to Netfilx’s marketing power and the other novel features of the production – such as publishing the whole series simultaneously.

Casting a big star with a track record of successful films is not a new idea, so choosing Kevin Spacey isn’t intrinsically revolutionary – nor is it unusual to use well-known directors or target genres that are known to be popular. To know for sure whether Big Data was important we have to know that Kevin Spacey was better than other actors who might have been cast in the role without the use of ‘big data’ statistical techniques.

This is clearly impossible, and perhaps an unreasonable demand. However we can get a feel for the Netflix approach by looking in more detail at the data they assemble and they way the use it.

Ian Bogost & Alexis Madrigal took apart the Netflix system for recommendations [7], reverse engineering it to see how it works – at least the parts of it that are visible on their website. Subsequently Netflix cooperated with them to explain some of their processes, granting some window into the company’s inner workings. In addition to the data they have about who rents which movies, they have have panels of expert reviewers who

…receive a 36-page training document that teaches them how to rate movies on their sexually suggestive content, goriness, romance levels, and even narrative elements like plot conclusiveness. They capture dozens of different movie attributes.

Using this data, films are classified into 76,897 “altgenres” that can then be used for recommendations (eg ‘African-American Crime Documentaries’). Presumably similar data was used to inform the production of House of Cards.

Altgenres frequently include the name of an actor or director – ‘Dramas Starring Sylvester Stallone’ or, more pertinently here, ‘Mysteries starring Raymond Burr’. Raymond Burr, star of 1950s TV Series Perry Mason, is the single most mentioned actor in the categories. Perry Mason director Christian I. Nyby II is the most mentioned director, and Barbara Hale, who starred alongside Raymond Burr is also very high, in fact directly above Clint Eastwood. Madrigal calls this “Perry Mason Mystery” – why does the system rank all things Perry Mason so highly? When questioned about the Perry Mason Mystery, Todd Yellin, the designer of the system, simply says ‘These ghosts in the machine are always going to be a by-product of the complexity’.

In a similar quirk, in a previous attempt I undertook to measure the importance of historical figures using data from Wikipedia revealed that Mircea Eliade, a virtually unknown Romanian historian, as the fifth most linked person on the site [8].

Clearly, it would not make sense for Netflix to cast Raymond Burr in House of Cards, even if he was alive. This highlights the obvious point that Netflix are at most using their data to guide their intuition about the production – they aren’t going make bizarre casting choices just because the system says. This is not step change from how TV production worked before, for example casting directors have always been guided by previous successes, perhaps now even actual statistics, to make their choices.

The T-Shirt manufacturer Solid Gold Bomb is a useful illustration of what can happen when algorithms are allowed to make creative decisions on their own. Their system advertised thousands of different T-Shirts with automatically generated slogans printed on them through Amazon – hoping to find success through weight of numbers. If someone bought one it would be printed on demand to avoid the expense of actually making such a variety T-Shirts, most of which would never be purchased. Unfortunately, their system automatically generated misogynistic slogans such as “Keep Calm and Hit Her”[9], among others, causing outrage and the removal of all their products from Amazon.

Big Data has been most successful in scenarios where occasional failures can be tolerated. For example we can accept a credit card being blocked if a fraud detection system suggests it has been compromised, as long as we can remove the block if the alert is wrong.

In the case of designing T-Shirts, society found the failure of the algorithm morally unacceptable, in the case of producing a TV show failure is too expensive: it’s impossible to imagining a commissioner defying a strong intuition and casting a seemingly inappropriate actor on the basis of statistical evidence.

Converging output

The success of House of Cards will ensure that next time Netflix look at their data, Kevin Spacey, Political Dramas and director David Fincher will seem even more popular. If they were blindly to go by the numbers, they might see that the best thing to make was another House of Cards.

This is a problem that advertising platforms, such as Google AdWords, also face. They want to show the adverts that have been clicked on most, because they are the ones that will most likely be clicked on in the future. However, if this was all they did then there would be no opportunity to expose new, potentially even more effective adverts to users – because a new advert will, by definition, never have been clicked on. This is described as the Multi Armed Bandit problem, and there are a number of mathematical solutions to it.

Another way to understand this problem is as it was posed by Richard Feynman [11]. He thought about the problem of choosing what to eat in a restaurant – should you have the dish you know you like, or try something else, which could be disappointing, or even better?

All of the mathematical solutions to this problem balance some amount of choosing the option currently thought to be best while occasionally trying out some riskier things. When applied to the Netflix problem, it suggests commissioning some shows that are very likely to be winners (House of Cards), but also to try a few riskier things which might prove successful, but whose success is not so well predicted by the data. In the cut-and-dry, high frequency world of online advertising, this might be a useful result. However, in terms of TV commissioning, isn’t that what already happens?


Social scientist Donald T. Campbell formulated the following adage which captures something important about the problems that Netflix and cultural producers more generally might face in formulating statistical forecasts of success

The more any quantitative social indicator (or even some qualitative indicator) is used for social decision-making, the more subject it will be to corruption pressures and the more apt it will be to distort and corrupt the social processes it is intended to monitor.

Economists have the same concept, due to Charles Goodhart [12]

Any observed statistical regularity will tend to collapse once pressure is placed upon it for control purposes.

Big hit Hollywood films already follow a well observed formula [13,14], which is frequently derided as mechanical. In this area, we might expect Big Data to allow a fine tuning of ‘safe bet’ mainstream productions – but surely not revolutionising the industries approach, since it’s already so mechanistic.

But in general, even if it is the case that House of Cards made important use of Big Data, it may not be the case that this approach can be continued into the future. Audiences, especially more discerning ones, might come to recognise the kinds of patterns it produces and become tired of them.

More importantly, if Big Data driven content becomes a regular occurrence writers will respond to this new context – they might parody or satarise the phenomena, choosing to deliberately ignore, exaggerate or lampoon certain effects to play on the audience’s expectations – in much the same way as the Spike Jones film Adaptation pokes fun at script writing courses.

The three reasons given above are intended to describe some limitations of big data in the creative industries. One reason why Big Data, as a slogan, has gained so much traction is because of its facility for very directly increasing revenue by improving advertising effectiveness. In this sphere it suffers much less from the problems above: a poorly targeted advert rarely causes offence, algorithms can optimise while preventing convergence, and there is much less reflexivity.

A cynic might think Netflix already knows this. Perhaps their real play is to understand their customers in order to serve better targeted adverts in their films, not to directly shape their productions.


What could big data do?

In describing how Netflix likely uses its data, as a guide to temper intuitions, a new horizon for big data in the creative process may have been opened. Rather than looking movie consumption data as a way of algorithmically generating TV programs, perhaps a more productive approach would be to think of the data as more grist to the writer’s mill.

In this case, why stop at Netflix internal data? Projects such as IBM’s trend forecasting look as though they provide data which is just as relevant. By crunching through social media data, IBM attempt to distill what’s capturing people’s imagination. In two published examples they predict in 2012 (incorrectly?) that 2014 will be the year of Steampunk [15], and more plausibly that cycling would be the flavour of the zeitgeist for 2014 [16]. Even more ambitiously, proposals have been mooted to try and simulate the entire social world [17], which, as well as sounding like the plot of a film, might augment a company’s ability to formulate novel creative output that keys into the collective psyche.

In some sectors, human input is a (prohibitively) expensive, error prone factor to be eliminated, but in the creative world it’s an intrinsic, and valued, part of the product. This does not mean that Big Data cannot play a part, but, unlike other applications, in the case it is to provide novel, inspirational support to the creative – and essentially human – process, rather than a substitute.
[1] Rajaraman, Anand, and Jeffrey David Ullman. Mining of massive datasets. Cambridge University Press, 2011.
[2] Narayanan, Arvind, and Vitaly Shmatikov. “How to break anonymity of the Netflix prize data set.” The University of Texas at Austin (2007).
[3] http://www.nytimes.com/2013/02/25/business/media/for-house-of-cards-using-big-data-to-guarantee-its-popularity.html?pagewanted=all&_r=0
[6] http://www.fastcodesign.com/1671893/the-secret-sauce-behind-netflixs-hit-house-of-cards-big-data
[7] http://www.theatlantic.com/technology/archive/2014/01/how-netflix-reverse-engineered-hollywood/282679/
[8] http://jimmytidey.co.uk/blog/michael-jackson-is-more-important-than-jesus-fact/
[9] http://www.ft.com/cms/s/2/3f78b4ec-8c40-11e2-8fcf-00144feabdc0.html#axzz37TqfDTZM
[10] Chakrabarti, Deepayan, et al. “Mortal multi-armed bandits.” Advances in Neural Information Processing Systems. 2009.
[11] Feynman, Richard P., Robert B. Leighton, and Matthew Sands. Exercises for the Feynman lectures on physics. Basic Books, 2014.
[12] Goodhart, Charles Albert Eric. Monetary Theory and Practice: The UK Experiencie. Macmillan Publishers Limited, 1984.
[13] Snyder, Blake. “Save the cat.” The Last Book on Screenwriting You’ll Ever Need. 1st edition. Ingram Pub Services (2005).
[14] Hauge, Michael. Writing screenplays that sell. A&C Black, 2011.
[15] http://asmarterplanet.com/blog/2013/01/22562.html?lnk=sa_steampunk
[16] http://www.ibm.com/smarterplanet/us/en/socialbusiness/blogframe/fff2012.html
[17] Paolucci, Mario, et al. “Towards a living earth simulator.” The European Physical Journal Special Topics 214.1 (2012): 77-108.


Last week I attended a week-long workshop in Beijing about the ‘cultural industries’  (The 5th International Doctoral Workshop on Cultural Industries).

From outside all you know are stories about extreme economic growth and Chinese tourists buying expensive handbags. During the workshop we were exposed to one of China’s anxieties: is Chinese contemporary culture OK? Is there enough of it? Why is it still to so fixed on the West?

I often got the sense of a blank canvass – which I found incredibly exciting. What if China is ambitious about creativity as it was about the Olympics?

Culture broad and narrow: the fact that the Chinese consumer still hankers after the design aesthetic of of Apple rather than a local alternative, but also in the narrow, classic,  sense of music, writing, film etc. China understands that it can claim the world’s most venerable heritage, its concern is to reanimate that inheritance.

I suggested that it was only a matter of time before the new middle class generated a richer artistic landscape and demanded alternatives to iPads specific to the Eastern market. More or less, I was told “We’ve been saying that forever!”. There’s no question that an iFetish runs deeper here than in the UK. Someone said they would never think of buying a Chinese designed car and noted the absence of native cultural celebrities.

The country’s vehement approach to progress of all kinds has lead to a tranche of academic research to find out what can be done, as well as a policy drive. We were told that China is building 100 new museums a year; sometimes without much to put in them (I couldn’t confirm this, it’s rather easy to suspend your critical faculties when it comes to the ‘China builds X new Ys every Z’ trope). Blair-era Cool Britainia was frequently mentioned as case study in cultural policy, apparently without any of the nails-down-a-blackboard rictus it connotes for me.

The connective theme that came out for me was the relationship between the market and culture – perhaps unsurprisingly given the topic and the fact that that two key speakers were an economist and a marketing professor.

Neo-liberal Cultures

There was a relatively uncritical attitude towards the market, especially perhaps in the presentations the students each gave – for the most part a tacit assumption was that cultural growth would be stimulated by an unfettered market, and that profit was a legitimate goal of cultural output. If national politics is a pendulum, it’s understandable that China’s momentum is to the right.

If you stood up at our own institution, the RCA, and spoke of money and profit and the free market in relation to culture in the way we heard in Beijing, you could expect a furious Q & A, at least from those who hadn’t walked out. More than an observation about Chinese culture, it reminded me of how easy it is not to stop questioning your own political context, although in the next section I do some work to defend the assumption that markets and cultural creativity have an ambiguous relationship.

Creative, Cultural, Industrial?

All of the speakers addressed the issue of how to define the ‘creative’ and the ‘cultural’, a prerequisite for talking about them; they are widely used in policy discussions. Intuitively we  kind of know what counts: books, theatre, music etc. But is Apple a creative, or cultural, company? Advertising? Type-setting?Defining  a meaningful subset of activity without encompassing everything, or nothing, and such that each member of the set can meaningfully be talked about collectively isn’t trivial.

Mostly, the answer was an extensional definition, using the  pornography: you know it when you see it principle. I didn’t find this especially intellectually satisfying. Given that ‘cultural’ and ‘creative’ seem roughly similar but both fraught with difficulties, for my own mental model I’m happy to abandon the creative industries as a category. It seems to me almost any job at all could encompass some creativity, and the more senior the position the more scope for coming to novel solutions. Taking the term literally, as ‘one who creates’ surely a plumber is more creative than an Arts Council executive? The answer that only certain kinds of creation count leads to the question of which ones, where, it seems to me, the answer is ‘cultural’.

So what are cultural activities? My first thought was that cultural outputs are ones that most directly address our higher mental faculties, reflection, understanding, aesthetics etc, as opposed to the more basic concerns: staying fed, watered and sheltered. Clearly though, the whole concept is fraught with value judgements. My time in Bristol came to mind, where the council constantly fought distribution of fliers and posters for club nights, but when an opera was staged the council suddenly found it acceptable to drape banner adverts all over the place. For them, contemporary music (for which Bristol is actually famous) did not deserve support, while classical music (for which Bristol will never be famous) did.

Maslow’s Heirachy was mentioned, and for me that captures the most important aspect of what any program of ‘cultural’ promotion might encompass. Some of David Throsby’s research into the “Work Preference Model” in Australia, which this I found absolutely fascinating, pointed in this direction. In a standard model, the more you pay people to do something, the more they will do it, unsurprisingly. Many people who consider themselves artists have to take on “portfolio careers”, working outside their artistic practice to pay the bills. But, as they are paid more for that work, in contradiction to the standard Work Preference model, they do less of it. Rather than become richer, they reduce the paid hours of work and focus on their practice.

Combined with discussed unwillingness to assign monetary value to cultural output (discussed in detail below) a case starts to build that money and culture are hard to reconcile. You might be able to use money to access the lower levels of Maslow’s hierarchy, but not the higher ones. In my view, these thoughts might offer some pointers on how cultural policy might proceed. If I’m arguing it’s not about the market, a natural thought is to turn to the government to support cultural output: here I’m stuck, because that doesn’t seem to be an answer either, taking us back to the wave of nausea that is Cool Britania.  My favourite cultural gestures are two-finger salutes to politics and the market.

David Throsby

David Throsby is a Australian economist who took us through various ways of measuring the economic benefit of cultural output. I love the analytic clarity of  economic thought. Firstly he looked at the microeconomic picture. Traditionally, economists use ‘revealed preference’ as an approach to discovering value. If you are willing to pay £5 for one thing, and £10 for another, then economists assume that the thing you will pay £10 is something you value more than the £5 thing. However, many people feel that cultural output does not obey this model – that even though people aren’t willing to pay very much for it, they actually value it highly. Strong evidence in support of this was presented. An answer comes in the form of the ‘stated preference’ approach. This methodology is a sophisticated version of asking people what they value in survey form (hence ‘stated’ rather than ‘revealed’). From outside of economics that might seem like a fairly innocuous step, from inside I suspect it looks more inflammatory. If economists can’t assume that monetary value is a good measure of actual value the whole discipline would unravel. It would be like the medical profession discovering that people preferred being unwell. None the less, when you have compelling evidence that the revealed preference method is not successfully capturing value then something has to give.

In regard of the macro picture he showed that cultural output is an is not well measure at the national level. The crux is that the standard national accounts, which every nation produces, do not include cultural output in the way that they do fisheries or mining. The answer is ‘satellite accounts’, which take the standard national accounts and attempt to synthesise a measure of the value of the cultural economy by adding up the output of publishers, theaters, the music industry etc. – categories that do exist in the national accounts. This approach is perhaps in tension with the the fact that much of the value of the cultural industries is not captured in any standard economic stat, as outlined in the micro-view. Perhaps unsurprisingly, these satellite accounts indicate that politicians probably undervalue the cultural sector.

Any economic account of culture is vulnerable to a interesting criticism: that the value of cultural is literally incomparable. Economists are normally trying to make comparative judgements, as in revealed preferences, often by the market. What’s the best change we could make to improve the situation? What’s better: A new train station, or better roads, or cleaner air?

Perhaps it’s simply a category error to try and address the question “better poetry, more police, or better health care”. David Throsby acknowledged this question during his talk.

Peter Stephenson-Wright

As a professor from a business school with a background in marketing, Peter gave us an insight into the world of the cultural mercenary: the advertising creative. Uniquely among the speakers he also addressed big data very directly; he made the point that creative decisions are likely in the near term to be guided by big data, but not made by algorithms, as I believe is often hinted at in the hype around data. As is often noted, Apple got bad results in focus groups about the iPad, but it still worked out: even if big data has more scale, it’s still not going to tell you which products will sink and which swim.

Peter also noted, in line with the research by David, that creative activity will not be straightforwardly motivated by money, job titles or bigger offices, and suggested that varied work was also an important motivator. He gave the example of rotating briefs between different teams in advertising companies, to keep everyone excited.

One remark which I found hard to ascent to concerned the work of Damien Hirst, and Peter’s former employer Charles Saatchi. He appeared to suggest that Damien Hirst’s status as an eminent artist was supported by the fact that he had made a lot of money for himself and Mr Saatchi. Further, he indicated, by presenting controversial works Hirst gave journalists something to write about, apparently making him an even better artist.

In fairness Peter did say that Hirst was also a good artists because people enjoyed seeing his work. An alternative story where Saatchi used his marketing savvy to inflate the value of unexceptional works, and that the role of journalists is not simply to fill their pages with manufactured scandal occurs to me – though perhaps that’s me being an idealist. I’m not attempting to evaluate Hirst’s actual artist value, just questioning whether making money for your patron has a bearing on it, or whether your output feeds journalists with a neatly defined controversy really matters. If making money is part of Hirst’s shtick fair enough, but it’s part of his output, not the measure of it.

What next?

China is an laboratory for what is possible in cultural ambition; if money, policy, scale or brute force can squeeze cultural vibrancy out of the ether then China is sure to make it work. But if there is some imponderable required, if culture can only exist as a descenting voice, as a rejection of market logic, then perhaps China’s cultural policy will be best captured by the symbolism of hundreds of understocked museums. Looking out from the train on the journey from Beijing to Xian at the endless tower blocks full of people with increasing disposable income and time, presumably hungry for excitement and novelty, it’s hard not to sense a vacuum waiting to be filled. I imagine some jazz-era-style, elicit musical explosion, a symbolic, non-specific rejection of the system in the manner of Beatles drug-tinged edginess or hiphop style life-on-the-streets audacity - only western analogues are all inappropriate because it will be its own thing with its own rules, and will have, of course, to negotiate the political landscape in which it finds itself.

And if that’s the picture for narrow culture, as I defined it, what about broad culture – where will China’s Apple come from? Perhaps it’s worth remembering that both Steve Jobs and Steve Wozniak were formed by the foment of San Francisco counterculture, not the output of a university program shaped by cultural policy.




This week doesn’t really exist because it’s mostly about coming back from Germany and going to China. Some small developments:

1) Considering writing a little sketch in D3 to help me Snderstand network algorithms

2) Started to get into the network literature by reading this: http://www.scottbot.net/HIAL/?p=6279

3) That post pointed me onwards to this:http://www.cs.cornell.edu/home/kleinber/networks-book/, which I have printed out and will read on holiday.

4) Bugs continue to come out of the mill for the Hounslow project.

Network Tiles

John Fass’s cork tile / rubber band / drawing pin methodology for eliciting social networks has become a focus of my work in terms of providing a context for people to discuss their communities, as well as the direct network information it generates.

However, the finer points of the methodology are not yet fixed, and there is also a question as to how robust the process is. Will two people, describing a similar network, generate similar results? If one person is asked to complete the process twice, will they produce similar results? Do these questions matter?

To address some of these questions, I took the opportunity of being on holiday with my family to get them to have a go. I asked them to fill out a network around where they lived, indicating “people, places and organisations”. I interviewed each person separately, half received more detailed instructions, half were shown a photo of a completed tile.

In terms of stability, I found the tile process quite variable. For example, my Mum & Dad have fairly similar networks, but produced very different tiles. My Sister and her two daughters also have overlapping networks, but also produced quite different output. This is in part down to the differing and (deliberately) vague instructions. Of course, it also represents the fact that different people will see the same things differently.

Overall, I felt that showing people a photo of completed tile didn’t help them understand the process very much. The single most helpful instruction (which I was omitting to start with), was to ask participants to put a drawing pin in the middle of the board to represent themselves. Although it is possible that someone might feel they are not the centre of their own network, making this instruction leading, it guides users past the intimidating blank canvass they are otherwise faced with.

Further thought is also required about the hypergraph aspect: should users be able to use one rubber band to link multiple pins?

Roland Burt Visualisations

Reading Neighbor Networks: Competitive Advantages Local and Personal, and looking at his formula for various measurements of the network (density, connectedness, access to structural holes), I was struck by how they could be explained visually in the style of Bret Victor. This may be a project for the future, it would be a great way for me to get my head around Burt’s work.

Continuing development of the Network Observatory

As ever, writing a web app is always a bigger project than you think. It’s the first time that someone else has tried to use the Network Observatory code, and it’s proving a challenge. For some reason, deploying the code to the Modulus hosting services proved to make it much slower, which is annoying, because part of the appeal of the Meteor framework is that it’s fast.

(Which reminds of me of another point – that writing for Meteor has meant a lot of difficulty in terms of not having a relational database, I’ve spent ages writing what is automatically handled by the Rails ORM. There must be a way round this)

Password reset emails weren’t working, the interface does unexpected things, and it doesn’t give enough feedback when users carry out actions.

Hopefully, I’ll tackle these and the app will start to become more mature and usable.

Art Vs Science

I’ve been working on an essay for ages, about the culture clash between the arts and the humanities, but it doesn’t seem to have taken much shape yet. For a long time I’ve been trying to fit it around the structure of an update on CP Snow’s Two Cultures book / lecture. It’s been something of a breakthrough to realise that this is not the correct starting point at all. A project to return to in the future.

What I’m Doing

I’ve wanted to update the headline version of my PhD for a while, but every time I do it, it turns out to be extremely time consuming. Now it’s done, I hope it will be a helpful way for me to describe my work to people; it’s also helped clarify things for me.


Tom Bryan came down to work on a new track for Rifff, our randomised-with-intent sequencer. As ever, plenty of bugs to work out of the system, especially since it hasn’t been used for so long (expired AWS keys etc). But I was pleased to get a whole, new, cogent track up — however at this stage the randomisation is perhaps too subtle to notice.

Barbican Digital Revolution

Went to see the Barbican Digital Revolution exhibition and was struck by how difficult it digital artefacts are to display. On the most pragmatic level, a lot of the exhibits weren’t working as intended, even Will.I.Am’s (‘Artist, Humanitarian, Philanthropist, Inventor’, according to the blurb) obviously very expensive installation had robot instruments that weren’t working. Durable, functional, robust digital / physical artefacts are hard.

Chris Milk’s large scale interactive projection, which, in its favour, was utterly robust, seemed out of place. As a piece of work, I think it’s scale made would have made it more suitable for integration with the architecture of a building rather than presenting it as another exhibit. I felt that might have been a way to minimise the emphasis on the technological aspect of it (‘Using interactive projection!’) and more on the effect, the aesthetics of it could stand on their own.

The most compelling exhibit was a video wall explaining the production process for the movie Gravity, which gave a really compelling description of what was going on, and how complex the production was. A technical feat, but with a purpose: to create the believable world that made a successful film.

Compare with the (musical) keyboard that had each of it’s keys mapped to radio stations from round the world. Cleverly done, well presented, but why? It’s an interesting idea, but I’m not sure anyone has to actually make it.

Roland Burt
Having spent a long time looking for an accademic backbone to my PhD, it feels like at least an area of the literature might be hoving into view. Previous forays into the literature around Human Computer Interaction have been informative, but since only a part of my project is about computer human interaction, it didn’t feel like it was foundational. More over, none of the various approaches to HCI seemed to particularly leap out as a powerful tool for designing better interfaces given my constraints.

The other path I was looking at was using social capital as a yard stick for the effectiveness of any intervention I might do. However, it’s such a broad concept, and it’s so hard to get hold of empirically, that it never really sat quite right.

Roland Burt’s work looks at social capital, but through the lens of networks:- what kinds of connections between people maximise social capital, performance at work, or knowledge sharing? This ties in neatly to the kind of analysis of online behaviour that I’m able to perform.

My hope is that his work, or work that I find through his, could become a grounding theme for my PhD.

This is my first week note, hope I manage to make this a habbit.

Ames gunstock

Ames Gunstock Lathe in the Science Museum’s Making of the Modern World exhibition

The Ames Gunstock Lathe is a tool for carving rifle gunstocks from wood. It functions by running a probe over an already shaped “template” gunstock. The probe is mechanically linked to a cutting head that produces an identical copy from a wooden blank.

According to geographer Jarred Diamond’s book Guns, Germs and Steel, the ability to make guns has shaped global history. Ian Morris, in his book Why The West Rules For Now echoes this sentiment, suggesting that mass-produced guns tipped the power balance away from nomadic tribes and in favour of the sedentary urban populations that we now take to be defining feature of civilisation. Mechanisms such as this lathe are clearly influential in the broad sweep of history.

Specifically, this tool was built in the Springfield Armoury in the United States. The facility’s ability to mass produce guns had a profound effect on American history, and is now a national monument and museum for this reason. The production techniques pioneered there also seeded the Industrial Revolution in the United States.

In terms of historical impact, this exhibit couldn’t have much better credentials for inclusion in a gallery about the making of the modern world. It was the novelty of the mechanism that caught my attention, but what set me thinking more deeply was the attached description:

“This machines’ legacy is the computer numerically controlled (CNC) machining systems that characterise mass-production today”.

Perhaps if the label had been written more recently it would have referenced 3D printing instead of CNC.

To me, it’s not clear the lathe warrants a place in the gallery on this basis. While superficially similar to a CNC lathe in terms of it’s ability to automatically produce a complex form, the two things are in fact profoundly different.

The authors of this description have not appreciated that the Ames Gunstock Lathe has no numerical or computational aspects at all.

The machine is so fascinating exactly because it operates without any level of abstraction. It takes as input one gunstock and makes another with no representational intermediate. In this sense it’s the absolute antithesis of the “information age” in which now live, as defined by the rise of abstract representation.

In fact the lineage that leads to modern computer technology and CNC tools was already well established by 1857. The Jacquard loom used holes punched in cards to control the patterns which it wove into fabrics, a genuine information technology. The link between the Jacquard loom and modern computing is unambiguous. The system of using holes in cards as an encoding method was prevalent in computing right up until the 1960s. Much of the standardisation of punch cards was undertaken by IBM, very much a link to the contemporary.

So the Ames Lathe, which was built 50 years later than the first Jacquard looms, doesn’t feature in the genealogy of CNC machines after all.

Disinheriting the Ames Lathe is more than just an exercise in taxonomy. Comparing the Jacquard loom to the lathe is a case study which can shed light on the defining characteristics of information technology.

Claude Shannon published A Mathematical Theory of Communication in 1948, giving an account of how measure information that is widely accepted. However what information actually is and how it is deployed in technology is less clear.

The Ames lathe is a vivid illustration of the contrast between highly malleable and liquid data which powers the modern world, and the non-representational physical object which has been so much less fertile in terms of innovation.

As far as I can think, the only functional modern device that users an analogous mechanism to the Ames Lathe is the machine used for copying keys at high street shops. Meanwhile, the informational approach of the Jacquard loom was already exhibiting the advantages that make information based manufacturing so powerful.

For example, the cards that controlled the Jacquard loom could be converted into electrical signals, sent over telegraph nearly instantaneously and recreated at some distant location. Conversely, by requiring a physical full scale wooden representation of a gunstock, the Ames lathe can only transmit a design at the same speed as any other medium-sized physical object.

Punch cards can be reordered to produce new patterns in woven cloth with very little effort, while for Ames lathe to produce a new design a whole new template must be hand made.

This ease of manipulation and transmission are the key features of information technology.

For me the inclusion of this lathe says more about the making of the modern world than many of the exhibits in the gallery that genuinely embody computer technology. By illustrating a technological cul-de-sac it throws into sharper contrast the path that progress has actually taken.

Balint Bolygo mechanical sculpture

Device using similar mechanism made by artist Balint Bolygo. In this image it is copying a cast of his head onto paper.

Does it matter if Twitter, defacto forum of online political discussion, is run as a private enterprise out of San Francisco? Could we do better than that?

I intend this post as a survey what might motivate a non-commercial Twitter clone, what it’s scope could be, and what other attempts have been made. It was triggered by at least three independent conversations that I’ve had about why such a thing ought to exist. It feels like everyone is thinking about it.

There’s a very productive analogy between political rights in physical space and those in digital space which can be used as a tool for examining this question.

In public physical spaces we expect to be allowed various political rights such as freedom of speech and the right to protest – for example on high street or in a park. It’s an intrinsic part of democracy, but it’s also part of an important idea that the space is owned by and run for the community.

Many people worry about spaces that are appear to be public but are legally considered private, for example in shopping malls or Olympic Park in East London. These ‘faux’ public spaces could be used to rob us of our freedoms without us noticing. At the same time, most reasonable people accept that you can’t hold a protest in a hospital ward or on a cricket pitch during a game. It’s a balance.

Using this analogy to prime our intuitions, have we got the balance between public and private space right online?

Facebook is an expensive thing to run, hosting billions of photos and messages; it constantly evolves. Without investment, it would not exist. And if you don’t like it, there are other ways to keep up with your friends. For those two reasons I think it’s at least reasonable that Facebook is a private enterprise.

Twitter is different, it’s a classic ‘faux’ public space. It is the public forum for digital debate. Several features ensure it’s a perfect fit for that role: public by default, the follower-followee model, and the @ message system which means that everyone at least has a chance of being heard by their target. From haranguing customer services to following the operation against Osama Bin Laden, it truly is the digital inheritor of the notion of the national town square.

Also, it’s extremely simple. They do huge amounts of work to fight spam and operate at massive scale, but the fundamental mechanism could be written on a napkin. This is not true of Facebook, or, for that matter of Google, or even Amazon. It could exist without lots of money to support it, much as Wikipedia does.

Twitter, or something like it, is the key part of a digital, participative democracy. For these reasons, the commercialisation of the public space that Twitter represents is a terrible deal for society. The balance is wrong.

Freedom of speech should not be at the mercy of a corporation’s terms and conditions. You should not have to see adverts to see what your politicians are saying.

So why not build a free, open version of Twitter?

If you built a Twitter clone only to provide a public online space it could actually be a bit simpler than Twitter itself.  No need for DMs or locked accounts, even favourites aren’t part of the core offer (nor are the complex system of “entities” that the site uses in the background). Just the ability to post and repost updates, to follow and block other users. That’s it.

The Twitter clone’s underlying API (the server) could exactly copy Twitter’s, bearing the previous scope limitations in mind. Just like Twitter, OAuth would solve authentication.

Then anyone could repurpose any Twitter clients they have hanging around to use the new service (I realise Twitter has banned using their API to build clients so not many people will have been developing them, but I’m sure some people have been messing around or have old ones.) A Twitter-style client is not a complex thing to build from scratch in any case.

Even if the API is conceptually simple, scaling is another issue. However, Twitter and Facebook have been open sourcing software that helps solve exactly this problem, for example Cassandra. Given all this, I wonder if building a limited Twitter clone might not be as hard as you’d at first think.

It would also be possible to build a federated version that shares the load between many server owners, though I’m not sure if it would be advisable, for reasons I’ll discuss in the context of Diaspora. If server prices became prohibitive, my first response would be to discard old messages – if the service did gain traction archive could be left to third party services.

Isn’t this Diaspora?

Diaspora is a federated, non-commercial version of Facebook. Fundamentally, it designed to provide privacy, as Facebook does, not the public space that Twitter does. However it does aspirations in terms of non-commercial ownership – in fact there is no single owner at all, instead it’s a connect network of servers owned by different people or groups (‘pods’).

This system points up the complexity of federation. Diaspora’s protocol has to be very complex to allow the transmission of messages between different ‘pods’ while ensuring only people with permission see the messages. One reason to do that is the legal protection offered by a service whose servers are not located in any one jurisdiction.

Our Twitter clone would alleviate a some of this complexity because all the messages are public, however I still think there is danger that it’s too conceptually difficult for average users.

I say that because on Diaspora the technical complexity is very visible to the user, and because of that I think it won’t catch on. Certainly it hasn’t so far.

In fact, I think Mark Zuckerberg saw this. Realising it was a fatally flawed Facebook alternative he donated money to the project to prevent a more threatening rival emerging.

What about App.net?

App.net is a service that provides an online identity. You can use that identity to log into any of their ecosystem of apps. The first app they offered was a Twitter clone, with the insanely confusing name of “Alpha” (they also have an app called Omega, both of which are presumably in beta). Their thought process seems to have been roughly similar to that outlined above, though focused not on notions of public and private space, rather on ownership of data.

My feeling is that App.net’s project is enormously broader in scope than an open Twitter clone, and that, perhaps for this reason, it doesn’t seem to have gained much traction.

What about encryption? What about the NSA?

All the messages are public, so there is no need for any attempt to hide anything. I’d consider this to be a considerable advantage since it makes the build so much simpler.

So what? 

I’d love to push it further, to a back of envelope calculation of what such a service might cost to run, perhaps restricted to just the UK.

Every day, every day. Every day on my laptop. If feel a bit like a prisoner in solitary confinement who forgets how to walk more than two paces: my arc of gaze is limited to the 13″ of my MacBook. It’s a voluntary arrangement, but it’s so useful I can’t get away from it. Will it be like this forever?

I’m interested in the iPod as activity-specific device. You can listen to music, but not browse the internet or send an email. As a result, it could never dominate your life like a laptop does.  I think we’ll see more of activity-specific form factors, instead of the ‘swiss army knife’, all purpose devices that pervade at the moment.

Steve Jobs said that Apple would not be releasing an e-book because “people don’t read“. Obviously some people read, what I take from that statement is that the e-book market is too small for Apple to bother with. (It’s seems like they were right: Amazon has a particular strategic interest in the Kindle, things like the Nook have not be very profitable.)

What Apple would rather sell is a universal device that can do everything, and therefor has a bigger market. The iPod was a beachhead, a personal device from a time when screen and processor tech made a multipurpose device impossible. Even then, the iPod targeted a use-case, listening to music, that is almost universal. As soon as it could, Apple bought us the iPhone and the iPad, which allow you listen to music, browse the web, any task someone can write an app for. This is a great place for them to be because the market is enormous.

Now they are stuck. What could make the iPhone or iPad or MacBook better? I would suggest there are essentially no improvements to be made to it (I’m not alone). The only things left are incremental tweaks to the OS, battery life, camera technology. Apple isn’t alone in this, phones and tablets all offer similar specs with few obvious areas for improvement, except perhaps battery life. Chromebook laptops are available at virtually disposable prices, and are increasingly reasonable offerings. Especially if you put Ubuntu on them.

The crux of it is that the tech to build a great device is not expensive or rare any more. A Raspberry Pi (£18) has (just) enough power to run an OS and a web browser, which is basically all you need. Any additional complexity can be shunted into the cloud.

The free availability of Android and tailored versions of linux obviously make a big difference, but perhaps the biggest factor is that we’ve stopped demanding faster and faster processors, there just aren’t any tasks a consumer wants to do that are pushing at this limit any more.

For these reasons I foresee that tablets and phones will be increasingly commodified (as do others) in the future. Probably laptops also, however for reasons I don’t understand no one seems to make laptops that are quite as good as Apple’s – perhaps because they have the whole area locked down with patents (just a guess).

I started by contrasting the universal device with the activity specific iPod. I think the pendulum might swing back to the activity specific device while the big manufactures are stuck in a cul-de-sac of increasingly commodified universal devices.

There are two reasons. Firstly, as devices get cheaper it will become feasible to own more of them. Secondly, the only significant improvements remaining to be made to devices are their physical interfaces, moving away from the “picture under glass” paradigm.

An example of this I’ve been toying with is the idea of portable device specifically for writing. It would have an excellent, real, tactile keyboard and a e-ink screen. It might connect to the Internet to save files, but would have no browser to avoid distractions. Without a backlit screen it could have great battery life and be very portable. It could be cheap, perhaps less than £100. I’d buy one.

I think this diversified future is something to look forward to. While Facebook and Google might still dominate the web landscape, perhaps in devices there will be a more pluralistic market. Lower barriers to entry and smaller markets to harbour niche manufactures.

Finally, I’d like to suggest this vision might be a more plausible frontline for the Internet of Things. At the moment, we mostly think of IoT as putting processing power in previously non-digital objects: often fridges, or smoke alarms, or bedside lamps. I’m not always sure these offerings quite ring true for me. Perhaps the slightly IoT-ified tablet or laptop will be the way that ubiquitous computing creeps into our lives. It seems more plausible the computational ubiquity will seep out through devices that look gradually less and less like a laptop, as opposed to leaping directly into the toaster or bicycle.


Every Big Picture event (run by Thought Menu) at Limewharf has been really different, all of them great fun. The last one was an opportunity for Vinay and Smari to sketch out their ideas for the future. Ed Geraghty also spoke, his thoughts were of a very different nature so I’ve not discussed them here.

Vinay & Smari’s visions were radical, and radically divergent. They only spoke for 15 minutes, so they didn’t get a chance to go into detail. Here are some of the salient parts of their views, as I understood them:

  • Smari (video here) wanted national governments entirely replaced by highly localised institutions running along “liquid democracy” lines, possibly using some of the software he has written. He also described a “craft economy”, which I assume means less mass manufacturing in line with his localism.
  • Vinay (can’t find video of this one) suggested that for humanity to flourish we must move manufacturing to the moon(!), while people on earth live in deindustrialised rural villages. Manufacturing on the moon is a necessity because of the dangers of nano and bio technology – in particular that (otherwise useful) biotechnology equipment could be used to engineer apocalyptic bioweapons.

Both of them described intriguing thought experiments addressing existential issues: climate change, disease, resource shortage.  But as I ran the experiments in my head – projecting myself into their hypothetical worlds –  I found myself in a rather disenfranchised society of limited political agency and narrowed horizons. They replaced our broken system with functional but politically uninspiring alternatives.

It’s very easy to conceptualise as yet unborn citizens as a kind of homogenised, infantilised mass whose biological needs are met by the paternalistic helping hand of the utopian envisioner. People are free to make the small decisions (local ones) while the big structural decisions have been taken for them, for example manufacturing safely sequestered to the moon like a box of matches hidden from a child.

I’m sure that neither Vinay or Smari intended this – perhaps it’s a consequence of the short amount of time they had to present. I wanted to point it out because I hear so many descriptions of the future that take an authoritarian approach to citizens.

Local democracy (liquid or otherwise) might empower people to take control of their village, but in so doing it must inhibit or discourage organising at the global or national level – because we all know that the design of a democratic process will influence the outcome, and here localism is the goal. Yet many of the problems we face are global. How can you reconcile the two? Implicitly it is because all of the biggest problems have already been sovled by the architect of the utopia.

In any case – who wants to live in a village? The idea of village life as idilic is an illusion of the rich. People move to the cities because of money, but also because of they are exciting. In their excellent book Poor Economics Abhijit Banerjee & Esther Duflo do a fantastic job of explaining just how boring villages are, and how important boredom is for population dynamics. They point out that in remote villages people will go without food so they can have a television just to ward off the boredom.  Cities are where ideas happen – why should we attempt to deny out future selves the excitement of urbanisation?

It’s striking how much Vinay & Smari would (I think!) dislike living in the worlds they propose. They both love the possibility of acting at the global level, and would hate the idea of living out someone else’s utopian blueprint – rightly so.

At all of the Big Picture events, we’ve focused on… big picture issues. It’s been about global issues, about the totality of the supply chain, about the possibility of large scale, leaderless cooperation. These things are a lot of fun to talk about.  Little Picture Days wouldn’t be so inspiring – how should we run the town hall? On what terms should we trade with the village next door?

It’s telling that the Big Picture Day events could only happen in a big city. How would you get 40 people with niche political views in one place if everyone lived in villages?

We should aim to address existential problems while taking into account that people, whenever they are alive, will be just like us. No doubt some will be happy to live a problem free bucolic existence, but many will want to live in a city and engage with the foment of ideas.

Some people, perhaps an increasing number, like the idea of a craft economy, but others get a buzz out of making the largest impact they can. Writing code scales naturally and globally, which is precisely why people like Vinay and Smri do it.

Our future selves will be as keen as we are to seek out challenges and sacrifice comfort for intellectual stimulation. It’s often said that the past is another country. In our heads the future is another planet – but we should remember that, at least for a very long time, it won’t be.