 Hello, good afternoon and thank you very much indeed for joining us at today's lunchtime lecture. Today we're having a Data as Culture lunchtime lecture. Data as Culture is the contemporary art programme for the Open Data Institute. I'm Hannah Redler and I'm the Associate Curator in Residence and I run the programme with Julie Freeman, who's an artist and the art associate. But we're really delighted to invite Giles Lane from Poboskis. We've known and collaborated with Giles for a very, very long time, way beyond the point where me or Julie started working with the Open Data Institute. And Giles is an artist who's been exploring the impact of digital technologies, network technologies, participatory practice and all manner of intersecting things for a very long time through his organisation Poboskis. The project I've invited him to speak to us about today is a project that completely changed my perception of what data is and might be and what it might mean to individuals. But I'm not going to give the game away. I hope that you'll join me in warmly welcoming Giles Lane. Thank you Hannah. Thank you very much for inviting me and it's very nice to see you all here. I'm going to talk today about a concept that I call data manifestation and in particular how I believe it can add a really significant dimension to the way we make meaning from data and I specifically mean digital data, digital data sets. As Hannah said, I run an organisation called Poboskis which is a non-profit artist led creative studios, how we've described it for quite some time. Essentially that means we're not really like other arts organisations and we're not really a consultancy and we're not really an agency with a little sort of a smersh of all of these different things. We have, over the last 22 years, since we founded, worked in a number of different disciplines. We've done an awful lot of collaborative work across different sectors whether collaborating with government, with lots and lots of different universities and different disciplines within academic fields. But also we do a lot of work at the community level, often bridging right the way across from highly localised and specific communities right up to larger entities like institutions, industries and governments. So I'm going to talk today, as I said, about data manifestation and just to give you a little background to this, it kind of emerged out of a long series of projects exploring the way that we use and conceive of emerging digital technologies and in specific, I'm referring to network technologies as part of that, in social life. So back in 2002, 2003, we set up a project called Urban Tapestries which was a very early attempt to think what the addition of geographic data to what we think of as content, so films, videos, text, photographs, how that would change the way that we would inhabit urban space in particular. That particular project, we built our own system, we collaborated with the Ordnance Survey to use a high-level mapping layer and we began to experiment with the very, very first smartphones which predated iPhones and what have you by about five years. And from that, we wanted to explore what actually happens when people are able to interact in a kind of live and dynamic way with the urban environment to be able to layer information and also access information that is there behind the scenes. So we did some experiments beyond that initial project. So over five years between 2004 and 2009, we had a research programme called Social Tapestries which expanded beyond the original project and took us into working with specific communities, housing communities, communities around public spaces. And we began to experiment with different forms of interaction with the environment. One of the projects we did quite early on was with the artist and engineer Natalie Jeremy Jenco and we had originally, back in 2001, commissioned her to produce a project she called Feral Robots and we adapted that and added it into the Urban Tapestries platform that we had created and we looked at what could people do if they were able to sense in their own environment different kinds of pollutants and then add that to the kinds of data that the national statistics or other local authorities could provide at a highly localised level and then add to that the local knowledge that people who lived in the area would be able to overlay. So we were kind of very much interested in this idea of data, both data created but also data sensed. And from that we began to kind of experiment with, at those days there wasn't really any open urban data. Essentially we were scraping PDF files on the national statistics website to get ward level and below data to add into these kinds of things. We did another project about a year or two later in 2007 called SNOUT where we created carnival costumes that had sensors but also they had displays on them as well and we did a sort of a mock carnival through the streets of Shoreditch engaging with thinking about what we could detect on the street but also relating that to community forums. There's a lot of community forums where people particularly in built up urban areas are concerned about the impact of pollution. Often these are in the poorest districts where there's all kinds of deprivation. So we were interested in what can we do, how can we explore these things, how can we make sense of the world around us. As we were doing these things, often what we find with digital technologies is that the experiences themselves are quite ephemeral. So we go along and we take part in a festival, we take part in some interactive element and the interaction is very momentary and you come away with it and you sort of have a memory of it but it's very hard to attack. We did some collaborations with some other research organisations and one of the things we noticed was where people are building interactive digital systems, for instance in zoos and museums, people would sort of wander around, they would collect things but nobody ever went back to the highly personalised web pages or websites of their visit that were created. So we thought this was kind of interesting and we started to think about this and in our own projects which involved digital interactions we started to make things. This is probably 2008, 2009. We did a project called Sensory Threads where we again designed a wearable set of sensors that a group of people went out and the sensors were sensitive to each other and as they went through the different sensors that each of the four people had detected different things about the urban environment and then relayed that to a piece of software which generated a live audio track. That was also uploaded to another system and that system was then replayed in a different format in an environment, a museum exhibition setting environment. One of the things we found was that the experience was interesting and people liked the sound but they wanted something more tangible. So as part of that we built a little printer that printed out a representation, a visualisation of their journey and the sense of readings that they encountered. We found this really simple little thing helped people recapture and remember their experience in a much more tangible way. From this we began thinking about what can we do more? How can we push this further? How can we take this idea of the tangible souvenir of an ephemeral digital experience and turn it into something much more profound? How can we start to think about the different data sets that we're beginning to interact with in a more profound way? That led me to think about the relationship we have to the world as individuals, as human beings. So the question that I would pose as I sort of slowly move into the project I'm going to focus on is how do we know? How do we know what we know? And how do we make meaning from what we know? And the answer to this is really quite simple. It's about the interaction of our senses with the various stimuli that come into them from the external world and how this is processed by our consciousness and then woven together with our memories and our emotions. This is what roots us to the world. It also helps us to make meaning of everything that we experience. So we began to think, could we begin to apply this to some of the data sets that we were interacting with? Could we perhaps investigate data in new ways? Not just visually as most data is represented but with all the senses that the human body has. So what I'd like to say about this concept of the human sensorium as separate to the way that we normally think of making sense of information is that it has a high complexity. It's a system of extremely high complexity and entanglement. If you think about these kinds of issues you think about the way that our different senses all interact with each other. They're not separate systems, they're part of a whole. It's very difficult to disentangle them and I think this is a key kind of concept. What it does that I think is extremely profound and has really interesting implications for our contemporary society and the way in which data driven systems based on algorithms are increasingly coming to make decisions for us is that this system of the human sensorium produces judgments that are based on multiple factors and also multiple dimensions that all work together. This is very, very different from the kind of way computer models that take data in from one and spit out a decision at the other end. I think it's qualitatively different. So, just as an example of that, think about... think about eating something. Think about eating something and you find it delicious. What is it that makes it delicious? Is it just the taste? It's also the smell, it might be the texture. It might be the way it feels in your head. It might be the way it feels in your mouth. Sometimes it might evoke a memory or it might produce an emotion. Occasionally, to add in a little bit of proust, it might even induce a momentary sense of disruption of your balance. This can then of course lead to all kinds of things which have nothing to do with the data. This is part of the human condition. My sense is, I don't think we are anywhere near the kinds of systems, computer systems, algorithms that could replicate that kind of complexity. And I think as long as we are designing systems which produce decisions for us that are not based on these kind of complex, entangled human judgment, we're really doing ourselves a disservice. So, there's another aspect to this which I think is also worth thinking about. And coming to this place and this space from the frame of art and being an artist, I would say that art and aesthetics are a very critical way to begin to think about this, this form of complexity. So for instance, when you encounter a work of art, it's the work itself, it's the experience of the work itself, sorry, that determines your aesthetic reaction to that work. Now that reaction, it could be one of awe, it could be one of delight, it could be revulsion or even indifference, but it is the encounter that determines your aesthetic experience. This is, of course, driven by very many complex factors as well. So, materials of the work, the lighting, colour, scale, your own memory, and your emotions. This is all part of the aesthetic experience. And there can be no right or wrong aesthetic experience, there's just difference. And in the realm of art history, the discussion around aesthetics is about the different qualities that different people perceive. So, there's no right, there's no wrong. I think this is a really important framing that we need to think about in terms of the way that we use data and the way that we understand data as well. From that, one of the things my sort of thesis is that much of the data that we're currently capturing through digital systems that's recorded, that's analysed, and then is used to make in inferential decision-making systems, it's principally driven from a visual perspective. So if you think about the spreadsheets, the programming, every way that the systems designers, the systems programmers are thinking about it is done from a visual interface. And we're constantly talking about visualising data, but we don't have many other ways of interacting with that data. You know, spreadsheets, graphs, animations, these are all the ways we're most familiar of thinking about analysing the data that are coming in. There are a few cases of sonic representations, and there have been passingly few, but there have been some rare examples of haptic interfaces. But on the whole, pretty much everything is based on a visual relationship. But of course, not all of our experiences are visual. And my proposition is that, as I said earlier, if we fail to encompass a more rounded, more inclusive set of the whole human sensorium, then we're really missing a great deal of potential in meaning-making. So, to dive in a bit further. Back in 2012, we were commissioned as part of a public art programme in Cambridge to collaborate with some scientists at Philips who have a small research lab in the Science Centre, the Science Park. We were asked by the scientists at Philips as part of this collaboration to think about a particular problem that we were concerned about. Philips is a big player in telehealth. Telehealth is a system or a set of platforms and technologies that are about embedding the homes of people with very serious illnesses with different kinds of sensors, relaying that information live to medical professionals and creating an interactive interface between them. So they were very heavily involved in this middleware area. And as of interest, they started to think about what would happen if we took some of those elements and applied them to people who aren't seriously ill, people who think of themselves as sort of nominally healthy, so probably all of us in this room. And there had been a burgeoning kind of industry growing up in the last couple of years before that. Obviously Fitbit, FuelBand, these different kinds of health tracking devices. But what we were kind of made privy to at that very beginning of 2012 and these statistics weren't generally known at the time, but something like between, I think, I'll try and remember the statistics, but essentially something in the order of 90% of all of the devices went silent after about a month. So these, expensive toys that people were buying, usually at Christmas and birthdays as people want to work off the little gup that they've acquired over the festive season. Because they weren't having the effect, people tended to, as usual, drop them in the back of a draw. This is well known with early technologies. But they were thinking, well, this is a real issue because if the personalised tools for gathering the data aren't engaging ordinary people, where is the possibility of any kind of future health benefit from actually gathering data about ourselves? So we began to think about this from the perspective of a kind of an artistic intervention. And our kind of first feeling about it was there's something really obvious here. There's something really obvious about what's going on, which is that how does a graph relate to the story of my life? If I'm looking at a graph every day, how do I connect to that? And I couldn't personally make any kind of connection to it. So we started to think about how do people mark change? How do they mark difference? How do they create mnemonic triggers for themselves that have meaning? And this is often done with objects. I mean, we're a highly tactile, material-based species. We like things. Just looking at the room here, as I always do when I give these talks, everybody is wearing something different. And you're wearing those things because you like the material, you like the feel of it on your skin, you like the way it looks. These are very different things to thinking about an abstract visualisation of the self. Humans are not, in many ways, always concerned with the abstract. We like things. We like stuff. So we began to think about different ways that we could mark the relationship that we have to the data that we might be collecting about our own lives and how we could begin to build a different kind of relationship to that since it was very, very clear that looking at a graph of how many steps you've walked or what your heartbeat was on a phone or a laptop simply wasn't attractive to people. So, as I've said, we saw this as being about the narrative of the self. How do you erupt data into the narrative of the self such that it has a kind of a tactility and a stickiness or a glue to it that is actually going to engage people with thinking about the relationship between what they do and where they want to be? So we began to look at the different things that people have, wedding rings, jewellery. But another thing we found is that there are an awful lot of people who have very, what appear to be quite random objects that hold significance and meaning for them. One of the classic ones I found was a pebble from a beach. I've met an awful number of men of a certain age who quite often, not every day, but perhaps on a regular basis carry a pebble on a beach and it's usually something to do with teaching a child swimming. But it acts as a marker and it takes them back to that place and it connects them to the narrative of who they are and who they want to be, who they think they are. So we began to think about how could we replicate that kind of level of mnemonic trigger, that level of meaning that people inject into a thing and carry around with them. And what we did was we built a series of simple data loggers that connected to a range of different sensors all off the shelf components. We collected, just within the team in the studio, it's about three or four people, we just collected a range of different data patterns over a week. Some of them were digital data, some of them were observed data as well. Step count was one, pulse rate, sleep patterns, blood pressure and other kinds of stress factors. And then we also mapped things like journey times for work or travelling to work. And then we began to think about how we could fold them all together and first of present that data in a way that it can interact with itself. So how we could begin to turn that into something that would create a tangible outcome at the end. So my colleague, Stefan Coopers, developed an algorithm that we could use. And we began thinking about how we would be able to flow a range of different data types into an algorithmic model and then from that generate some kind of shape in three dimensions. At that time we were thinking very much around 3D printing and access to 3D printers. So this was also partly thinking about a model for what might happen say 10 or 15 years into the future. And in the same way that the desktop publishing revolution of the late 1980s eventually fed through to the fact almost everybody has an inkjet printer at home or access to one very quickly. I think things like 3D printers are quite likely to be extremely accessible within 10 to 15 years. So we got a little stuck around thinking about the form that this 3D printed object derived from digital data would take and we began to look to nature for inspiration. And at that moment we realised that the maths and the geometries of shells would offer us an extraordinary complex range of expression of data in a 3D format. So what I'm now going to show is just a few of the examples that we generated from this very early project. So we collected, I'm trying to remember how many shells we did. I think we printed over 50 shells or generated over 50 shell models. Initially it all printed in standard sort of 3D printed in standard laser centred nylon. And then over time we selected a number of them and explored working in different materials. So these ones, this is silver. This one's been printed in glazed ceramic. There's one in glass and there's a couple in various different polymers. And we were interested also in the nature of scale. The scale of an object, how big does it have to be or how small does it have to be to have meaning and what does the scale or the size of the object do in terms of its affordance as a meaning-making object. And then also the different qualities of materials. So all these things have a very different feeling to them and they make us think and respond and reflect to the object in different ways. What I'm going to do beyond this is actually let me just move to the next one. So you can see images of one of the shells, in fact this one here that was printed in silver. So you can see along the z-axis is the average pulse. This was I think about a week's worth of data. I think this is even my data. The length of the shell is determined by the average pulse rate of my heart. The surface pattern scaling, so the depth of the ridges on the surface is determined by the quality and length of sleep. And then here you can see the growth segments is controlled by the number of steps. So again how much I did or didn't walk over a week would affect that. We did some measures of anxiety to think about growth disturbances. So whether or not it feels... Of course there's another element to this which is printing in slightly different materials gives you different surface resolutions. So the silver tends to be smoother but if you printed it in say a more... a material where you can have greater resolution you'd see more difference there. And finally we also did some measuring of air pollution. And this affects the scaling and the length of disturbances. Yes. We did a... We used a... Yeah. Yeah, a sweat. The name of the sensors has just gone out of my... Yes. GSV, Galvanic Skin Response. Yeah, so we used that to kind of take a measure of anxiety. I mean the point was we were using off the shelf things rather than looking at highly customized sensors. And that's one of the ways that we often work is to look at what's available and work rather than get too detailed and access things which are too far from the ordinary public. Essentially most of what we do is about inspiring other people to kind of replicate our work and build on it. Yes. Is there any extra environmental data? Did you only do the pollution or did you like to other... other environmental data like temperature, humidity? Not for this one, not for this one. Yeah, sure. So for this particular project we didn't... we didn't test other... we only had a small range of sensors which are the ones I mentioned before, sleep, pattern, GSV. But none of the others. I'll just go through the next bit. So what I'm going to talk about is some work we've done more recently with Birkbeck College. We've been collaborating with a colleague of mine and collaborator for many years, Professor George Roussos. And George is one of the academics leading a research project into Parkinson's disease and they're designing a very simple way for people with Parkinson's to collect information about their experience of Parkinson's. And Stefan who designed the algorithms for these is currently doing a PhD there so we've been applying some of the work we did before with this particular project. So this project is called Cloud UDPRS if you want to go and look it up. And as I said it's about mapping the wide variation of symptoms that people who have Parkinson's experience. If you don't know Parkinson's is a disease which encompasses a high variability of symptoms and you'd think this ought to imply a high personalisation of medical care and the management regimes for their illness. But, and this is the key thing, care management is determined upon a scale called the universal Parkinson's disease rating scale. This is derived from around about 70 different factors of motor performance against which an individual Parkinson's patient is classified against. Now these factors themselves are collapsed into a single summary statistic so you know a 1 to 100 number and this is used to then assign treatment for patients according to where they lie on the scale. Now this can be a very effective mechanism to communicate the multiple dimensions of Parkinson's disease as a linear scale of progression. So obviously people near a 100 have got a much higher incidence of Parkinson's symptoms and people lower down have got a lower incidence. But because of the enormous variation it's common for people with very very different symptoms to score similarly on the scale and yet they require very different treatment and care. Now one of the things that we've been looking at in a particular way of allying data manifestation with the Birkbeck's working kind of analysing the rating scale is to think about how important it could be in the future for people to express the uniqueness of an individual's experience of a very complex illness like Parkinson's. It would be very important for the individual in terms of personal dignity but also it would be extremely useful as a tool to convey the wide variations in an easily appreciable manner to policy makers. So this is one of the reasons where we think it's more than just kind of interesting pretty stuff. It actually could have a really direct function. So what I'm going to show you here I'm afraid the image isn't terribly good but I've got a few others. What I'm going to show you here is four shells that we generated about three months ago. From four individuals who have taken part in this trial and contributed their data to this and what these four shells do is they show you just taking three data sources they show the extreme variability on people who I think from my understanding are relatively similar on the UDPRS scale. I believe so. Something like that, it would be something like that. I don't have that information because that would be too much data for me to have because this is obviously personal by data. But the relationships you can see I've got the four examples printed at a slightly larger scale here and by all means when we finish come up and have a feel. Essentially what we did was we mapped the tremor in left hand to the length of the shell spiral so long length or compressed. The overall growth of the shell i.e. its volume is related or mapped to their left leg agility and one of the other tests which is called two target finger tapping again on the left hand side with left hand side data for this particular trial this is mapped to the frequency of the ridges on the outside of the shells. So you can begin to see for people who are scoring similarly on the scale that's going to assign them treatment there's actually high variability in the difference of their experience. You think that's just three out of 70 different factors that could be analysed in this way. So we can begin to add many many more data sets into this and add increasing complexity to the growth patterns of the shells things like the number and increment of rotations ripples in the curve and the sweep twists additional twists could be used and then we can also begin to look at adding additional geometric entities onto the shell surface so shell surface deformation could also be driven by additional data sets that could be bumps, nodules, spines those kinds of things. As I said also we only dealt with some three data sources from the left hand side of a patient but because there is a significant difference between the left and right hand side of the body we could begin in future to think about quoting what you might think of as a clam shell so two shells together to look at the difference between say what a patient experiences on their left and on their right hand side So with just three things we've been able to demonstrate a really high differentiation and individuation between four different people So as I said you know the thing to think is well how useful is this what can we really do with this kind of technique My feeling is we're living at a time where there is so much data we're swimming in ever increasing amounts of data from all different sources whether it's automatic sensing data from systems and products and services around us to the data we're generating from our activities and our bodies How do we make sense of this we're constantly being told that it's overwhelming us but is that because we're mainly using a very very narrow band of our sensory capabilities to actually try to make sense of this stuff and if we begin to expand it out we can look at other ways of stimulating the senses there's at least 11 major senses that the average human being has sometimes people think of that up to 19 and I heard as much as 28 but if we begin to think about how we can use a much more complex arrangement of these senses then I think there's an enormous amount of potential in rethinking what data is what it means to us and how we can begin to make sense of it in the future I'll stop for that Hello Hello Great presentation Giles I wasn't quite clear at the end when you were talking about the final thing you were talking about with the four shells that you degenerated Do those Would those have any function or any role in the medical process in terms of like say the doctors like supposing each patient had a shell and the doctor picked up the shell for the patient that was like sitting in front of them in their consulting office and sort of could get a kind of idea from the shell of that the sort of base condition of their patient Are you thinking of some role like that that could be directly used for? One of the difficulties that a lot of people for instance with Parkinson's face when they say or someone says I've got Parkinson's what does that mean and what does it mean to other people so having a way one of the things we've found is people having something they can actually show and say well my experience of this disease feels like this and it's qualitatively different to what another person with Parkinson's might feel so there's that element the ability to communicate some sense of what the feeling of a condition is there's another aspect to that again that we would like to look at but it's quite tricky which would be is there stuff being collected in the data that because we're looking at it only from a visual sense in terms of the graph and the pattern of visualising the data collected from somebody with an illness are there patterns that we're missing because we're only looking at it from a visual perspective so this is one of the questions that we would like to ask in a future project exploring with clinicians to think about might there be ways to perceive different patterns because if we're only looking at data on a screen perhaps there are things we're missing it's very hard to tell then again you've also got to be thinking about what you're collecting in the first place but yes there's another aspect also to the issue in terms of having something that you could take to a physician and have a discussion with them about your condition so I mean we were imagining excuse me I'm going to sneeze in a minute excuse me we were imagining that you might produce these every month or every three months essentially curate an exhibition of your experience whatever it was whether it was just your general health or something more specific and that could then be you'd be able to over time see the changes actually made manifest in an object so that becomes a way of thinking about how the long term patterns which are very hard for us to perceive particularly with digital technologies it's all about the now moment and the speed of progression of things we don't have very many digital technologies that are particularly good at showing long term patterns or revealing long term patterns though I don't see why we couldn't actually deliver that but having something which you could look to over a period of time could be a very interesting way of thinking about the relationship between yourself and your experience there's one other little aspect that I'm going to mention beyond that which we began to think about as we began exploring this idea we began to talk about this idea and present it I presented this at a data ethics symposium at the Alan Turing Institute a few months ago there were quite a few cyber security people there and people involved in medical ethics and one of the things that people got very excited about is this idea that you could carry highly sensitive medical data around with you in a form that nobody else would understand other than you and a physician who you would be able to have a discussion with to interpret the object so we're talking about this idea of data veiling that you could veil you could carry your information with you in a fairly secure format because it would be almost impossible to reverse engineer back to the original data that generated any of these objects and that could be another kind of profound way of changing our relationship to data I just wanted to because you're doing something that I'm very fascinated with biofeedback with how we relate to our environment and how we relate to ourselves so I'm wondering if there's any scope would there be any scope with using that data as a live feedback to the body of the patient so I mean if my hand is trembling I'm realising for a lot of Parkinson people we actually not realise that it's trembling they have a way of getting a feedback from that from that data that goes your hand is trembling be aware of it I mean not with this because this is a slow process because obviously we have to collect the data first then we have to clean the data and then it has to be fed through into the this project is called Lifestream so the Lifestreams process so it's not a dynamic thing it's more about in a sense creating a snapshot of a period of time and often this is data that's collected over a week or a month what you might like to do is look at the project which is called Cloud UDPRS and I think it's UDPRS.net but we can sort out the link because they've developed Buckbeck and their partners have developed a smartphone app which I think it does not quite 20 of the factors so we're using data from that so you can test yourself and that's where the data has been derived from for this set of shells alright so you use that app to collect the data and that's an experimental app to actually give people with Parkinson's the ability to measure themselves along some I think slightly less than 20 of the 70 main factors used in the rating scale alright that's great thank you you're welcome are there any other questions for Giles? well on here no you said at one point about when people are using Fitbits and stuff like that there's all this data being gathered but people don't often return to the place to look at their data and I think it made me consider why a lot of those organisations have failed on that point and often those big those devices are made because they want to cull vast amounts of data from the big population to use either in medical research or in social to understand how we work and what to sell us and so I think that maybe that's why that gap they're not actually interested in us improving our own health perhaps you're right when back in 2012 when we started this project when we were told unofficially that I think around about 90% of devices after a short time like a month or so were basically going offline the first question was how do they know and the obvious thing was the devices reported back to the manufacturers now they didn't reveal this till 2013 halfway through the year these devices have already been on the market for a couple of years so essentially if you bought a Fitbit or any of those kinds of devices they were reporting back to the manufacturers your data but they didn't bother to tell anybody that they were doing this and I think as you say that throws up enormous questions around why the devices were being made for whom, for whose benefit and again I think it presents real issues in terms of thinking well are we getting the devices we really want or are really useful to us or are we actually sleepwalking into essentially being producers of data for somebody else's benefit and essentially I think by one of the roles of artists is often to challenge and question this thing I mean we're often at the forefront of kind of hacking together and building things that don't really do what they're intended to do and a phrase that I rather like in this is creative misuse I like to think that what we often we're doing is we look at the the technologies, the tools that are being provided you know they're being made and manufactured because someone wants to make a profit somewhere and that's not necessarily a bad thing that's just an observation of the world around us people that's how our world operates but I think we've got to question these things all the time and there is room and space in our society to challenge and to play with these things and to disrupt the flows that otherwise I think begin to really hem people in and tramle us in very tight ways so for me this isn't really about creating products that people would sell I mean lots of people say oh I'd love to have a shell it would be so nice but actually that's not really the point the point is to think about how do we expand our senses how do we explore the range of our capabilities and our capacities to interact with the world interact with the systems that we're creating in the world and feel that we have agency at the end of the day all the work I'm interested about is about feeling empowered it's about feeling you have agency and that you're not locked into somebody else's system this is of course extremely appropriate given our current situation socially and politically but I think this is the role that artists offer not all artists but people who work in this kind of way this is still a visual representation of data it's a tactile one as well you can come and feel them can you give some idea of what order of... in which order, which other senses could you use and which one would be the next ones to try well if you think about the way that we interact with the world everything is, if you can see everything is visual there's very little that is entirely invisible there are some things obviously smell is something that we don't tend to see but one of the aspects about a three-dimensional object is we don't just feel it and we don't just look at it we also have a spatio perceptual relationship to it so that's to do with our kinetic sense and our balance sense of balance and proprioception which is the relationship we have to things around us which tell us where we are in the world so all of these things no matter how big or small they are once they become an object they operate on multiple sensory levels in a way that a visualisation on a screen simply doesn't you don't have a spatial relationship to the data on a screen you're not able to use your sense of proprioception you can't really use... you can't have a tactile relationship to the data if it's represented on a screen these are all the ways in which I think this does something quite qualitatively different there are a few other people over the years who have experimented with smell and combining different chemicals to produce smells which create a sensory environment that you can relate to different forms of information I think somebody even produced a smell generator quite recently which could be hooked up to different data sets I can't remember the name of the project off hand but I think these are beginning to be ways in which we can think of... what would it be like to be able to move through an environment perhaps where it might be a room that we decide to fill with different capabilities so you change for instance air pressure years ago I did a project where we were trying to... it was about abnormal... the difference between normal and abnormal brain development and we were trying to give people who were able-bodied a sense of what it might be like to have their senses disrupted so we had hot and cold zones we had quadrophonic sound which changed your sense of... and we also put the floor on a slant so you moved into an environment where your normal capacity and it was dark that was the other thing so you couldn't really see and there was a range of different ways that we disrupted people's senses interestingly enough an awful lot of people found it a really comforting environment to be in which was the opposite of what we anticipated but I think if you can begin to... think about the complexity of the way our bodies are able to interpret enormous amounts of stimuli coming in from the outside it gives us an incredibly rich way of thinking about if we're overwhelmed with data then perhaps we just need to put the data into another form that we can interact with it on a much more complex way and not be afraid of having to collapse everything down into what we call summary statistics On that note Giles, thank you very much indeed Thank you everyone