top of page

Mixing Artwork & Private Data - Elliott Round.







I spoke to Elliott Round the co-founder of M-XR to discuss what he has been finding out about data privacy collection in addition to how he has been turning his data into artwork.


"The vast majority of my artistic work has always been under the genre of 'conceptual art', that is to say, I am more interested in what I am trying to portray than the visuals outcome itself. For some time now I have become unstuck with such a concept of which I find interesting and instead have continued to explore the world of digital art, expanding and progressing my skills within this field."


How did you get into film & VR (Virtual Reality)?


So I studied film at university and whilst I did that I did a lot of special effects, computer graphics, green screen stuff so that I could add extra stuff into the video content that I was making. Before and after finishing university I was working at a couple of effects companies, the second one I was working at started dabbling with VR. Prior to that, I didn't have much understanding of programming and the gaming side of things, it wasn't something I initially thought I'd be interested in, but when I got to know it I could see the potential for other things outside of gaming because it is much more to do with building a world than it is to do with content and I think we focused a lot currently on the content side of things but we haven't given much attention to places we can go and visit.


How can we protect ourselves when it comes to our data privacy?


Besides from just using certain things less and also using your browser in incognito mode, there are a few other third-party browsers available which do offer complete browser privacy, and also be mindful of your cookies and cache because they can record some of your data.


What do you think about the future of data collection?


Data collection itself isn't bad it's just what you do with is, I'm hoping we're gonna start moving towards more ethical means as opposed to pushing for capitalism because then you can get some quite interesting and helpful things out of it, especially in the field of medicine and also for mental health, thinking along those lines could be a benefit.



"Advertisement companies using big-data as a means to predict human behaviour. "

I have been working in the field of emerging technology now for the past 6+ years, having followed the growth and application of Artificial Intelligence and Machine Learning closely. An area within this domain that became incredibly interesting to me, from an ethical perspective, is that of advertisement companies using big data as a means to predict human behaviour. This accuracy and precision of these algorithms are quite hard to fathom, and often gets easily dismissed when it is explained to people outside of the field.


I would argue that we as people don't even fully understand ourselves that well, so for someone to state that a computer can predict our behaviour and interests sounds like bullshit - "We are too complex to predict", "People are so unpredictable", "you don't know what is going on within someone's head" - are common rebuttals to that statement. Yet the reality is... they can, and they do. People's hubris, unfortunately, gets in the way of them being able to comprehend such a statement by putting themselves (humans) up high on a plinth as some complex intricate beings - with enough data, and examples we can be easily predicted and behaviours can be identified.

When we hear about corporations having data on us, it's often the case that people believe this data is just categoric, factual things like our gender, where we live, genres of music we like - similar to facts on a TopTrumps card, is often why some individuals don't particularly care about their data privacy as they are not giving anything "too personal" away... it is just facts that they are putting out into the world. However all of the data that is recorded on us is in the form of Time-Series data, this is information that changes and grows over time. For example, rather than just logging the Top 5 retailers that I shop online at, that information is recorded with a timestamp, allowing much more complex information to be derived from this - showing when we shop at which store over the course of many years and the frequency.


Data like this is much more valuable for predictions as it is easy to start identifying trends, similar to looking at a graph of the planet's heat and seeing how over many years it is on the rise - we can, therefore, predict with reasonable accuracy of what the heat may be in the year 2025. The same can be true for our shopping example, we can identify that this individual is moving away from stores which are high street brands, and moving onto high-end fashion slowly, this would be a point whereby we can see a shift in this person's wealth and desires.


This is not what all the person does, it's just things like shopping behaviour. The type of data that is logged is far more detailed, and each of these is known as 'data-points', such as:

  • Where we are, when our phone has used public wifi in a shop, restaurant or the transport network, for how long and when this occurred.

  • The things we search on google.

  • The things we watch on youtube.

  • The music we listen to on Spotify, the genres and the mood.

  • Statuses we type on social platforms, and whether we post that status or delete it - these are all logged.

  • Our scroll rates on different websites and media platforms to see what posts are engaging to us or did we scroll on.


"Why is it that Tinder & Hinge may only be advertised to a newly single person, rather than someone who is in a long term relationship."

The list can go on and on, the frightening thing about a lot of this time series data is that they are not separate from each other and this is how much more personal information can be derived. Why is it that Tinder & Hinge may only be advertised to a newly single person, rather than someone who is in a long term relationship - it is due to multiple channels of time series data showing very personal things, as behaviours suddenly change across all of these channels.


Besides, we are not just looking at your data, but every person on the planet. This means that whilst I may not know that someone has just become recently single, we do know the behaviour of 1000+ other people who have, ten or so of these people may have similar behaviour to our person of interest - we can then very easily start to predict their next steps, which videos they might watch and will the music tastes change as a result - whether the algorithm guesses correctly or incorrectly, by the fact it attempts to predict the next steps will allow it to self improve, but fine-tuning its settings meaning future predictions will be far more accurate.


Asides from just predicting behaviour, if it can be predicted then it equally can be persuaded by changing those recommended videos on youtube, shows on Netflix or even the playlist suggestion on Spotify can all be used to alter our behaviour over a long period of time, to gain the desired outcome. This all sounds incredibly long-winded for just selling a product, but the reality is it is far more efficient than just showing random people adverts until one of them, maybe decides to buy.


This is something I have wanted to illustrate for a long time, without knowing how. This has led me to the realisation, that due to my age (25), the majority of my life has been via a device and therefore nearly all of my relationships with friends with loved ones has been captured through some form of a telecommunication system, all the way from meeting a person, the relationship with them and whether we have lost touch or are still close friends today, all of my interests over time as I've grown up to now, the music I like, the films I watch and my daily habits and routines.


Whether I like it or not, this has all been tracked one way or another. I saw this as a perfect opportunity to start expressing how all of this big-data is recorded on us, by using my own data to create a sort of AutoBiography of my life - without myself having to log any of this information (just data that is already out there). By compiling this information into such a visual autobiography, the final piece would not only be detailed, but highly personal to myself and in some senses rather too private for me to even want to share on the internet - yet it is that breach in my own privacy which is what I want to make others aware of, as if I can show the entire history of my life, the good and the bad, then it is very apparent that yours is out there too.



Losing Touch With Someone.

A Parent.

A Long Romantic Relationship With Someone.


A Work Partner.

A Three Year Friendship (A)

A Three Year Friendship (B)

Comments


bottom of page