Interview: Mat Dryhurst

“Hang on – I’m sorry to interrupt, but your face is reflecting and glowing on my glasses and it looks so cool. One second.” There’s the loud snap of a screengrab and a roar of laughter from the other side of the Skype call – and another, quick sorry. Then, between rolling cigarettes, Mat Dryhurst scrolls and types as quickly as his ideas rattle out of his head.

As a music producer, he’s released solo records on the PAN label and worked on projects with its founder, Bill Kouligas, but he’s also a fervent student of the Internet. In collaborating with theorists, web developers, philosophers and audio-visual artists, he explores how sound, image and text can be used to understand and further explore how we exist online.

As the growth of interest in his long-term collaboration with Holly Herndon gathers pace, he’s also launched a self-hosting media platform, Saga: a website that he’s conceptualised, coded, and now runs as a response to what he sees as the exploitation of creative labor online; cutting out third party servers and allowing artists to host, track and manipulate their work. In conversation, he divulges his motives in creating Saga, how the Internet has mutated for artists, and what Saga could potentially achieve.

Okay, to start: has the development of Saga been a reaction to the problems that Soundcloud and its users have been facing, or has this been in development for a while and these problems have just bolstered your resolve?

It’s something that we’ve been thinking about for a while, largely thanks to the #TheIndyWeb discussion. Amber Case and Ben Verdmueller are two people who I follow closely. They began discussing an issue within computer science that’s become super relevant to my work right now: the problems of publishing your information through a centralised third party.

I’ve also been thinking about issues of compensation, because my wife Holly [Herndon] and I do this for a living. The issues with Soundcloud have emerged as a parallel discussion, but the symptoms of it are much the same as those of the #TheIndyWeb: that if you end up publishing your work through someone else, they have control over that work and you’re pretty powerless as to what they end up doing with it.

Why do you think the issues with Soundcloud have seen such a strong reaction from its users and the press?

If you were to judge Soundcloud purely on providing a service, I think that they’ve done a pretty good job. I was working in Berlin when they moved to set up the company there, and their initial strategy to curry favor with the electronic music community seemed quite transparent.

Soundcloud approached people who [find it] perhaps easier to adopt new things, and more interested in the social aspects of sharing music. But if people who are independently-minded start pushing their work toward these large warehouses of information, these warehouses will eventually close their doors to your concerns because they have to make their money somehow. Since that community seems to be the first people that were hit by this shift in practice, one of the things I feel that the kick back has presented is that people perhaps feel a bit betrayed, maybe.

The idea of uploading a sound file and hosting it on a player that everybody can use has become standardized in a way that seemed foreign five or six years ago, but you also don’t – or shouldn’t – have to go through these platforms anymore in order to do that. Soundcloud provided a convenient service that had social features. Now, I don’t think that they’re providing anything that couldn’t be easily replaced by an independent alternative.

I don’t blame Soundcloud for taking the steps that they are taking, but with Saga, we don’t give a fuck about everybody. My loyalty is to independent artists and if you can really focus on the needs of a small community and don’t have to necessarily take money to do so, you don’t have to pretend to be everything to everybody.

What’s struck me about how this conversation has evolved is that the issues typically brought up in arguments against a digital upload onto a centralized, third party site are economic ones concerning ownership. What’s more complex, though, is when it’s less explicitly about economic ownership and more about the specific context of and an audience for the work. When an artist raises concerns of their work being spread online, the reaction tends to be “When you makes something, it’s out in the world and there’s little you can do.” Is something is that Saga is reacting to?

I’ve been in touch with DJ Sprinkles while writing the Saga code. A friend of mine told me that she was presenting something in LA, and came back to me and said, “Oh shit – you’re basically talking about the same thing.” I got in touch via email and asked, “Could anything that I’m building right now work with your own ideas?” It’s very easy to only really look at your work after you release it, and particularly if you are working from a position of privilege, or general ignorance.

If you want to publish a track and have as many people as possible hear it, and you’re not concerned about misinterpretation because prior to release there’s nothing there to misinterpret, that’s fine. But although DJ Sprinkles and I come from different musical legacies and are active in different communities, we have musical and philosophical commonalities that mean that the work stands for something. We are concerned about being misrepresented.

We have this idea of “maintaining ownership” in order to milk as much “value” out of the work as possible. For me, it’s less about that. In fact – and I think DJ Sprinkles would agree with me on this – I don’t care if x number of people share this. But I do care about where it’s shared and how it’s shared, and I also care about who is making money from this.

Especially if you’re not really making any money…

Yeah, and it’s also just the principle. I’ve been working within independent record labels since I was 18-years old. I started out working for Crass. That was a place and time where a community who were very without commercial means figured out a way to distinguish the work, and the ways in which people received that work. There is a fatalistic approach toward things today: where people say, “We put this out, and then this shitty thing happens to it, but that’s just how it is.”

I know. As DIY communities and not major labels, why should it be exploited in such close circles?

Exactly. I’m not concerned about discussions surrounding big companies who own master recordings and want to lock down the Internet. It seems unreasonable and I actually don’t want to use an Internet that is so controlled. A lot of the tracking systems that Saga uses are tools that YouTube have – but YouTube doesn’t give you access to them. When you watch an Ellie Goulding video on YouTube in Paris, the site sends you an advert that is specific to your location and to what else you’ve watched in the past, say, four days.

They have the ability to serve you contextually relevant information, but they don’t give that to their users because that’s how YouTube makes their value. This is my realist position – if the artists themselves had those tools, it could lead to some cool creative possibilities that a small group of people might want to explore: “Hey, I know you, you’re my audience.”

How exactly are you able to use that technology for Saga, and what would you like people to be able to absorb from it?

This is when using my new terminology will be important. There’s a new project called Ethereum that uses the Blockchain, which is basically a big spreadsheet that programmers use to write what they call “smart contracts”: the programmer’s attempt to cut out the middle man from any kind of transaction.

There are two levels. One: self-host your shit. It’s up to me in a way to make that easier for people, which admittedly I’m still working on. And two: get used to this logic of being able to assert yourself in each individual case.

In the instance of working directly with a platform, say, an artist could say, “Lauren, you can post this track of mine on your website and you’re allowed to host it until it accrues 10,000 plays. After that, I want it to disappear from your website.” You’re able to provide that framework for them via the smart contract.

I feel that the smart contract model is going to become pretty standard in the next five to ten years. Ethereum have raised something like $20m dollars and are meeting with Richard Branson on private islands. They’re calling it “Web 3.0,” “The Decentralized Internet.” For me, it’s all about introducing this logic of getting rid of third party intermediaries, because I think they’ll become obsolete in a few years.

For my own practice, it’s also interesting to know who is out there. We’re so used to making records and pushing them out towards this assumed audience, but we have smarter tools now. If you have a Twitter account, you know exactly who is sharing your track, or article, or film. You develop an awareness that becomes standard practice. That’s exciting because when you start talking about the potential counter-cultural opportunities for releasing music, understanding, identifying and speaking to whoever else is out there is cool.

The closest analogy I can come up with that in regards to Saga is that if someone hosts your material on their page, they’re basically giving you that part of their page. I’d hope that people would act responsibly, because there’s definitely prank potential in that, but what you do with that opens up so many different possibilities for creative opportunities. I’m really trying to get people past this concept that “when you release something into the world, it disappears.” It seems so at odds with how we actually engage online now.

If I came to you as an artist, how would I use Saga’s back-end interface to track who posts my material and where? Am I able to reach out directly to those people? And if so, on what parameters am I really allowed to do that as an individual? The idea of online privacy has become so emboldened and complex: that when you move into and between different online realms, and a stranger approaches you about something you’ve done on what you see as your private Internet, that can be seen as invasive, not inviting.

Absolutely. That brings up a really big point that I’ve been thinking about implementing in version two, because right now Saga has privacy issues in the way that the code is written. It’s funny: I worked on privacy issues with a group in San Francisco five years ago. It’s something I actually care about. But I’ve built this thing that’s tracking every IP address. We get way more information than even we as developers need. It’s kind of a problem.

For version two, I’d like to pick an arbitrary number of plays for any location, and you won’t learn where this work is being posted until it’s triggered that number of plays. That’s one of the more elegant ways that I can think of addressing the problem of your regular-ass person posting something to Tumblr being “exposed” for doing so. But it’s hard. I have friends who actually are in exile over this shit – who vehemently dislike the idea of tracking anything online – but my opinion on it is that until legal protections come into place, tracking isn’t going away. And I feel far more comfortable with regular, creative people having access to these tools, and having some kind of parity of access with large corporations, than I do with just the large corporations having that access.

So, your pragmatic position is that if tracking was a more free and open process, it would be less tracking as targeting and more tracking as a potential conversation?

Oh, for sure. It comes down to saying, “In music, tracking isn’t really a conversation,” but if the first instance of this can precipitate that conversation within the music community, then it would be really nice to work through that and figure out where our comfort level with it all is. My personal opinion about privacy is being worked through with Saga. The thing for me – and as you put it, it’s a very pragmatic position – is that if we just pretend this doesn’t exist in the way we publish our work, then we’re kind of leaving room to be taken advantage of.

But it’s also difficult because, from an ease–of–use perspective, there’s a lot of information on Saga to sift through if you put out a “popular” video. It is kind of like an NSA approach, where you’re just inundated with everything. It’s a totally valid criticism and I think that I’m going to make a call on it for version two, but I trust people in this community to use it responsibility way, way more than I trust YouTube to.

The introductory blurb on Saga opens with the statement that “Work is currently being exploited online.” “Exploited” is a loaded word. Who or what is being exploited, and by whom, and what can you realistically do about it with Saga?

I do feel that the amount of labor and time that goes into creating work online has been severely undervalued by those who derive “value” from it. There were a bunch of people working from the public resource of the Internet who made empires off the back of free labor, and it takes places precisely in the locations where the creative work is often the last thing to be considered and compensated. I see that as a principal issue of exploitation.

I think that the Internet has been this wonderful and exciting thing, but also that a bunch of people basically convinced everybody to just give away their shit and it’s turned everything on its head. Much of the decentralized web is about very early Internet politics: self-hosting and controlling material, where things look different from different people. That’s one of the big principles for Saga. I’m like, “Why should everything be put through one standard Soundcloud template? Why the fuck is my work beholden to that?”

I think listening to music on Soundcloud is the opposite to that of reading a book. When you’re reading a book, and you see there are fewer pages to go, you can derive enjoyment from that because you’re getting to a physical point in the text-as-work that subtly informs how you feel during the act of reading. With Soundcloud, it’s movement with the joy and intrigue stripped out. It doesn’t tell me anything about the work – just the model that it’s hosted on.

Totally. Soundcloud solved an interesting problem at the time, but they basically built a waveform delivery service. That tells me that their focus and concentration wasn’t necessarily on a quality of experience. To that I would say, try and give an identity to the work other than hosting it in an online warehouse.

On the one hand, you could opportunistically see it as a means to be more successful with your work, but on the other hand it’s just like all the work that people produce: you want it exist within a distinct environment that in turn contributes to the work. When it comes to electronic music producers, generally their work goes further because they start at a club night, or a radio show, or have incredible artwork, or are involved within a hyper-specific community of people. All of these factors distinguish the work, shape it and make it mean something.

There’s this super-weird “poptimist” logic that goes, “Yeah, Arca! Your obscure electronic music video will be covered on the same page as a Rihanna video! That’s great!” People will be checking out the Rihanna thing, and as a result they might check out Arca. People will disagree with me, but I see this as a flattening. It may have contributed to some people seeing your work, but you lose something in the process. I’m like, “Actually, Arca’s very, very, very different to Rihanna.” She’s representing completely different things and when you flatten these things, you lose opportunities.

I was on the Overground in London recently and I heard an announcement that started “Could all customers please…” I thought, “When did I stop being a passenger and start being a customer?” I’m wondering what your take on the listener as a consumer is, particularly since you’re not using the tools you have with Saga to create targeted ad revenue?

That’s a cool point. I think it speaks to this general flattening. The idea of the “user” is an issue that comes up constantly in the tech world. We have this strange approximation of things. When I first started to do research for Saga, and was looking other, larger data collection companies and the tools that they were selling to people, one of the most interesting elements was how language has changed in advertising communities. It used to be an abstract audience. We didn’t know who was turning up where. Then the audience became demographics. That was their way to quantify people, and you can see how demographics have really, minutely altered.

Urban Outfitters is a product of demographic thinking. “We’re going to place these stores next to coffee shops in Brooklyn and in Shoreditch.” These companies have now gotten to a point of precision where demographics have shaped individuals: people are approximated as archetypes, but the groups are getting smaller and smaller.

Have you seen the hand-drawn memes of the sad, lonely guy, surrounded by all the niche club culture tropes?

They’re amazing.

That’s the perfect example of that to me. Plenty of people share and laugh at them, but it’s an incredibly niche thing to relate to and find funny.

It’s true. This social minutia has always existed, though. It’s not like if you push against that system you aren’t going to find hyper-specific things. In fact, I encourage that process. I think it’s wonderful that these new archetypes are being formed.

The thing for me is that when subcultures existed pre-internet, or during a nascent period of at least the “social web,” you would have local distinctions between things. It meant something different to be in Glasgow than it did in London. When you look at online life, there are these little islands – these little pirate ships of individual groups that float around everywhere – and the world that’s being built around them is very standardized and weirdly penetrable.

You see this with discussion groups on Facebook, for example. A lot of really smart friends of mine who are discussing some fucking awesome things are using Facebook to do so – sharing links and information, contributing to the knowledge-base of Facebook – because there’s not a better alternative. And when you extend that analogy to music, I’m like, “Actually, we’re lucky.” For the kind of music that you or I care about, there are so many people within this community with technical aptitude. Holly and I have noticed that a shit-ton of developers are coming to our shows, for example. There’s a huge Venn diagram going on there.

When you start building your own stuff you can then start to build walls, if you want to. You can also start to embellish this in ways that are perhaps imperceptible or hostile towards people, or in ways that are really open to people, but it’s your choice to do so. There’s not this huge leap between doing that and also protecting your interests – or at least just manifesting your interests online – that also leads to a far more cornucopic online existence. Have you seen the “Boring Dystopia” Facebook group, actually?

Yes. I love the painted traffic lanes that allow you to walk on the pavement faster. Faster to do what, exactly – spend more money?

Exactly. But the extra level of meta-irony is that it’s a Facebook group, so you have people spending all their time posting these things and following the “Boring Dystopia” feed. I posted an image to “Boring Dystopia” of a white collar dude in a coffee shop browsing “Boring Dystopia.” We’re reaching a moment when we’re revelling in the transgressive potential of complaining. The next step is to actually just build something else.

A core tenant of these platforms is that they were built for integration with other, similar platforms. Is Saga a destination site, where you can only experience the content if you know the URL or specifically what to search for, or is it going to be embeddable and shareable as, say, YouTube and Soundcloud are?

This is where I start to become a bit of an antagonist. In its most basic form, Saga is an embed delivery system. Right now, you can share work hosted on Saga pretty much everywhere. There are a few places where it doesn’t work, but I actually want it to not work even more on those few places.

Is it because of what those few places are, or purely on the idea that you like it not working?

I like it not working everywhere. It’s both, and it’s connected, and for what I feel are good reasons. Saga doesn’t work super well on Facebook partly because it’s janky, but also because I don’t want it to. Facebook likes to take your work. They have a whole business model around it. When you post a video to Facebook, Facebook caches it locally so that it can load faster for people because they have their own incentives to do so, and they use really tricky, cheeky ways to do that.

I’m building something that so, when you post a Saga link to Facebook, it resolves as an image that tells you why you shouldn’t be checking shit out on Facebook. You can change that image if you want, but I really like the idea of using the means of how you publish to kind of ostracize certain things.

Facebook is like a closed ecosystem designed to extract value from things, and in my mind it would hurt people to not be able post a link Facebook, because that’s how people consume a lot of stuff. But it almost is a “statement.” If you imagine a scenario where a large group of people decided that they wanted to use that logic, it would potentially present a problem for Facebook. In terms of posting the embed code anywhere else, though, it resolves pretty well on Tumblr and Wordpress, and so on. Twitter is something I need to work on a little bit more, because it’s a bit weird. It doesn’t work as seamlessly as YouTube yet, but there’s no reason why it couldn’t.

You talk about tailoring an audience, which I think is something that’s been fuelled by the YouTube generation: if you can reach everyone, then why wouldn’t you want to?

Perhaps there’s merit to this concept of, “I want to reach as many people as possible because I want to take down as many barriers as possible.” In my mind, the jury is still out as to whether that helps you. This whole idea that not posting your material to Facebook is somehow limiting the number of people who can see it is just not accurate. There are all of these principles that could be manifested in the ways in we publish our work. But you’re not going to end up doing that right now, with these services that resolve beautifully everywhere.

The whole argument with Soundcloud is that it’s so easy, but playlist culture has penetrated everything: this idea of seeing this Amnesia Scanner song as having the same value as that Nicki Minaj song. It doesn’t, actually. People say that YouTube is this democratic environment in which anyone could stumble upon your video, but not just anyone stumbles upon a video. VEVO has these insane deals in which you get served particular VEVO videos depending on your age group and other factors. They have their own hierarchies in place.

Going back, you said that work is able communicate different things to different people in different locations. You’re using the example of precise geo-targeting for advertising, in tandem with the platforms like YouTube. But if we’re taking the advertising out, should a work communicate different things in different locations? And if so, why?

Two things. Firstly, I’m again removing myself from that dialectic because I’m not here to tell people what to do. Secondly, though, I do find it very interesting. You prefaced this with the issue of academic language being non-accessible. I’ve done this, everyone’s done this: “Whoa, I have to have read ten things just to be able to participate in this conversation.” I think it’s really important that philosophers can talk like philosophers, but there’s inefficiency to that process if you’re looking to make a connection with people.

I did a big commission for German National Radio earlier this year (which contractually I can’t post online, which is a shame. I worked my ass off for it and it just disappeared). I did two radio plays with the idea of experimenting with data mining as raw material for a work. One was called “Mine” – a short thing that’s been posted to Soundcloud, ironically – and the other was called “Muster.” For both, I went through exhaustive, very human-driven data mining procedures to develop research on exactly who is going to be listening to this work.

The story, music and everything about the work were created in response to what I knew about those people. I looked closely at the Facebook page for the event and tailored the event specifically to what information that I could discern from the people that clicked “Attending.” I also presented this experiment at UCLA. The target audience didn’t know what I was doing in advance and when I surveyed them afterwards, the majority were like, “There was something about this that felt familiar, but I couldn’t put my finger on it.” I was like, “It’s because the Chinese restaurant that I referenced in the story was the same one that’s opposite your old high school.”

This is definitely in that antagonistic artist role, but using this logic of data collection practices as a creative tool and to see how it affects advertising processes opens up this entirely new category for art. It made me think, “Can you challenge people in more precise ways?” The whole idea of producing an obscure, multi-genre club track in 2015: is that really the most contemporary way to challenge people, through some anonymous club aesthetic? This category of making work is new, and it’s also kind of terrifying, but it doesn’t have to be opportunistic and it doesn’t have to be populist. The best thing that can come from experiments like that is to make people really aware that a lot of the information that they’re receiving is targeted in such a way.

How much of these projects is about a guy sitting in his basement making a program for other people to use, and in exactly whichever ways that they wish to use them, and how much of it is “Mat Dryhurst’s Saga: a pet project for his own personal fascinations, trying to work through his own grievances with the Internet”? And do you anticipate a point where you will really have to let go of what you want out of it, and instead tailor it to what other people ask of you?

Oh, great question. In around 2011, I started this art project where I built a patch that would survey everything I was doing online and make music from it. I did it because Bill [Kouligas] had invited me to present something for the PAN festival in New York. I had a pretty punishing day job at the time, so I wanted a process that was functional and communicated the issue at hand: that I was too overworked to make art in the first place. I put together a record with the material but once I listened to it, I realised that the aesthetic of it had already been presented with Holly’s own record, Platform. I was bored of the idea, and other people in the community had started using this intimate, clash-y, “concrete” aesthetic for their sound. So I canned it.

People make records that talk about imagining a new world, but I thought, “I’m in a position to actually make a gesture that could impact this.” Something that people could use. Saga, then, is my “response album.” I want people to use it not as a product but as a project that is tailored towards them. I don’t see a distinction between my coding and my music. It’s all creative work to me. And the art world has a problem with that: if artwork does something, it’s almost disqualified from being an artwork and that is insane to me.

If you make a Tweet that changes the game, which people have done, that – in my mind – is as powerful as a song. And if you make a piece of software that changes the game, that’s as powerful as an album. It’s all basically smart people associated with a community sharing things with each other, and I don’t discriminate. In a way, that’s a pretty dangerous way of working, right?