The Invisible Hand – part 2

I promised more on the Blast Theory workshop I attended a couple of weeks back, and here it is. The two days were kicked off by Matt Adams of Blast Theory explaining why they’d titled the event with a term the Adam Smith had coined a couple of hundred years ago. Smith of course wrote about the Invisible Hand as a good thing, turning selfish acts into selfless ones:

By preferring the support of domestic to that of foreign industry, he intends only his own security; and by directing that industry in such a manner as its produce may be of the greatest value, he intends only his own gain, and he is in this, as in many other cases, led by an invisible hand to promote an end which was no part of his intention. Nor is it always the worse for the society that it was not part of it. By pursuing his own interest he frequently promotes that of the society more effectually than when he really intends to promote it.

Adam Smith The Wealth of Nations Book IV, chapter II, paragraph IX

However the Invisible Hand that we examined during the workshop had the power to be good, a more sinister. We were looking an profiling and personalisation. Matt started with off with an example of profiling. Mosaic is Experian’s database of UK households and organisations use to identify markets. My own day-job at the National Trust uses Mosaic, but you don’t have to be an organisation to access at least some of the data, their ForSite app is available for free  for your iOS or Android device. At the National Trust, we know that a number of the sixty segments into which Mosaic divides the population are more likely to become members than others, so certain postcodes might, for example, respond more positively to a fundraising appeal. Of course its a pretty broad brush, and describes a population by the people they live among, not as individuals. Its all about inference, not fact.

But technology enables organisations with the inclination to collect more and more data on individuals, not just addresses, and modern manufacturing processes allows a degree of mass-customization that moves the world of profiling into one of personalization. Matt spoke of the chain Zara, the high street face of Spanish company Inditex where staff collect feedback every day from customers, and send it to HQ, where designers collate and respond to what customers are looking for within weeks. Apparently only 3 or 4 units of each design are shipped to branches, and using the daily feedback from stores and a  2-3 week production process, the entire stock of a shop will change every 11 days.

Matt’s question to us all was, what does this mean for the arts?

Of course personalized art has been around for decades. Matt cited  Allan Kaprow’s 1959 work, 18 happenings in 6 parts; the neo-futurists’ ongoing Too Much Light Makes the Baby Go Blind (which is not just personalized but also randomized – a roll of a dice defines your ticket price); and Lev Manovitch’s Soft Cinema as examples.

Blast Theory themselves are not alone in the contemporary application of technology to create art: Invisible FlockNon Zero OneConey; and Urban Angel all create worlds that are part theatre, part game experiences that create unique experiences for each audience member. Many of these experiences collect data on their participants, is this a good thing or an abuse of privilege?

In the discussion, concern was expressed about the power of profiling and personalization. I found myself playing devils advocate. Yes, profiling is a blunt tool, and can be used to exercise power unfairly upon minorities, but personalization, if its done right, is a dialogue. Some years ago, when I was a young man, I liked Levi’s 620 jeans. Then, with some retro rock and roll, Levis introduced 501s and 620s disappeared from the shops. I was gutted, I’m not a handsome man, but I’ve got a pretty fine set of pins, which 620s showed off at their best (or did at least, when I was slimmer), tight in all the right places (around the calf I meant!!). 501s were baggy, which might have been fashionable, but what I would have given for the manufacturing process that dangles now so temptingly in front of us now, where those few of us that were not persuaded to convert to 501s might still buy 620s, even better than they were before because they’re made to measure.

Of course the debate turned to privacy, such a vital topic in these post-Snowden times, but even here, despite my abhorrence of the state spying on its democratic masters, I found myself an apologist for Big Brother. Is “Privacy” actually an aberration, a blip in the long history of society, brought about only a few generations ago by scraping together enough coin to live in the luxury of separate bedrooms? For most of human history we’ve lived with a different concept of privacy, where communities knew pretty much everything that when on, and privilege bought some people only a reduced number of people to share the toilet with.

I think we concluded that the debate was less about privacy than power. We discussed the politics of knowledge, touching upon Marilyn Strathern’s simple hierarchy of data, information and knowledge. Data being the unprocessed stuff that we see (or sense by any other means); information being that data organised in some way so we can begin to see some sense in it; and, knowledge is information that has an effect, that makes us change our behaviors in some way. The fear is that data is being harvested  by large, wealthy organisations, in such huge quantities and the “common man” or woman can not make sense of it. So we let these organisations organize it into information, and offer us knowledge that manipulates our behaviours to the organisations’ benefit. A new Invisible Hand that isn’t working for the benefit of society as a whole.

Kelly Page offered us examples. Doubleclick was a company that monitored websurfing, following clickthroughs on banner ads especially, a form of anonymous behavioural profiling. They acquired an off-line catalogue company called Abacus with the intention of merging their anonymous data with the  personal information within Abacus’ database. (though after Microsoft called foul, they were prevented from doing so by the US FTC. Later they (and their data) were acquired by Google. Page worked for a time at DunnHumby, who run Tesco’s Clubcard, and surprised us with the revelation that Inland revenue use club card data to validate your tax return.

But why should big and sometimes anonymous corporations be the arbiters of data, information and knowledge? We explored ways in which artists might take the mechanics of big data and transform it into different, playful information and knowledge. Blast Theory shared projects that tried to do just that from early experiments with Prof Tanda’s Guess-A-Where to a more recent and on-going collaboration with National Theatre Wales. Then John McGrath and Katherine Jewkes both from National Theatre Wales came on to talk about being an organisation that started out as an on-line community, and how even they struggle to cope with the “small data” that their participants have shared.  But they also gave us a taste of a lovely game/perfomance that imagines having to smuggle yourself of the border of a newly independent Wales. Part of that game involves creating a fake passport for yourself, and they were surprised by participants willingness to give their real names and data rather than making stuff up.

Giles Lane of Proboscis told us about a fascinating co-creation with Anglia Ruskin University and the R&D section of Phillips who have a interest in telehealth. Their idea which echoed our earlier discussion around the politics of knowledge, was to take an individual’s medical data and turn it into a 3D printed model, a talisman, a lucky pebble, which they carry with them. A beautiful object used as a tangible locus of meaning, mindfulness, rather than the dull data which means not enough to change behavior.

So can art use big data/infomation/knowledge to benefit society? And what are the ethics of doing so? That was the meat of our discussion, and Blast Theory, having collected our thoughts (data), are organizing them (infomation) and will shortly share them in some useful format (knowledge).

When they do, I’ll share a link here.