On Reverberations of Feedback Loops

Nicolay Boyadjiev in conversation with Anastasia Kubrak

Today, any inhabitant of a city is treated as a computational user by default. Smooth interfaces and real-time applications augment our urban experiences, while subjecting us to processes of profiling, quantification, optimization, and isolation. Cities start to resemble cybernetic feedback loops between inhabitants and their habitat: while we consume space through technology, the space consumes and produces us. So what does it mean to be a human user in today’s technological infrastructures? What are the strategies that enable users to gain more agency and be repoliticized? Anastasia Kubrak, a designer and researcher, sat down with architect and design strategist Nicolay Boyadjiev to discuss the role architecture and platform urbanism plays in empowering users in urban environments.

Anastasia Kubrak (AK) | Could you tell me a little bit about your background? What is the relation between your interest in platform economy and your architectural work?

Nicolay Boyadjiev (NB) | I’m an architect, design strategist, and researcher currently based in Moscow where I’m the design tutor of The New Normal program at the Strelka Institute for Media Architecture and Design. In broad terms, I’m currently working on various projects centered on platform urbanism and the spatial causes and effects of data and cognitive extraction at city scale. Since 2016 I’m also one of the co-authors of GoogleUrbanism, an ongoing media project and speculative proposal focused on the spatial component of platform capitalism and the attention economy.

AK | Today’s citizens are increasingly treated as consumers of politics. Do you think the user is an extension of a consumer, or are they different?

NB |The user is made; by definition, this is a position created in relation to something. With this in mind I like the idea of the user as designer. Your actions, whether you want them to or not, make something happen in relation to something else. By being tracked, monitored, or just by being, you have agency or are given agency because you are doing something. To give a blunt example, if you pay with your credit card inside a store, you may be revealing your purchase history and personal predispositions to a credit agency, which in aggregate may lead to the re-organization of the store layout to incentivize further purchases. In that sense, you could claim that by being a consumer you are also the ambient co-designer of the store interior. Even though you’re consuming experience, your activity is also producing something, and by doing one thing instead of something else, you have operationalized something. This is different than the position of a consumer.

Citizen, Consumer, User, 2018. Image by Anastasia Kubrak

Citizen, Consumer, User, 2018. Image by Anastasia Kubrak

Within the context of politics, I think there are many current narratives that make apparent the rise of a new type of authoritarianism. If you look at millennials in the United States, for example, there is a growing number on both sides of the political spectrum who are disillusioned with what we still refer to as neoliberal democracy, bringing about new sets of responses and expectations. One version of the politics of the user is a position where, as users internalize the fact that they are being monitored and their passive actions carry agency, there is an expectation that the system should know by default what you want instead of users making conscious decisions. As opposed to actively making a choice, you know that you’re being recorded, and therefore by passively being yourself you’re somehow showing your values; this should lead to political action. Moving past citizens as consumers, this version of citizens as users is problematic, and I think we should advocate against it. Even beyond the NSA or Google, decentralized “trustless” platforms have similar tendencies.

AK | What is the key difference between user and citizen in contractual terms? There is a presumably democratic social contract between a citizen and the state in the form of a constitution, but it is very different for the platform. Are you less empowered as a user?

NB | From the very origin of the term, citizenship always implies some sort of dynamics of exclusion. In the original polis, citizens were only a small minority while the rest were slaves. User, on the other hand, is an open concept: everything that can be registered, recognized or addressed by a platform is a user of that platform. In that sense you could think of it as much more democratic, as it includes groups, composites, and objects, although again I think the term itself needs updating. The categorical logic of addressability, or the politics of terms and agreements are definitely different from the social contract of a constitution, and so are its dynamics of partial or temporary exclusion. As an architect I have a limited perspective on this, and I would be careful in making claims about whether the user is more or less empowered, politically speaking. But what is clear is that the politics of platforms are raising new sets of questions and concerns around citizenship that we are only now starting to think about, and that we’ll continue to grapple with as design questions for probably decades to come.

The difference is that with the state, in principle everyone who’s part of it is theoretically under its gaze, authority, protection, and so on. The terms of belonging, borders, and exclusivity are fairly straightforward in most cases: you’re either inside or outside of the state. With platforms, what is interesting is that you could simultaneously be seen by some, but not others, at various points in time and according to different contexts, to the point where exclusivity itself is fluid, dynamic, and contextual in relation to many other factors. There’s no border or boundary that you’re crossing, but rather “access” or “registry” that you are given; the line serves as an enabled/disabled link between nodes, rather than as an inside/outside division. Since in theory anything can be addressed, the exclusion of a user from a platform is different from the exclusion of a citizen from the state. And platforms are responsible for various conscious or accidental conditions of exclusivity. This is a broader topic much more systematically addressed by Benjamin Bratton in The Stack (2016), but with platforms we are witnessing not the dissolution but rather the multiplication of borders, each with their own terms of partial or total exclusivity.

Citizen vs. User, 2018. Image by Anastasia Kubrak

Citizen vs. User, 2018. Image by Anastasia Kubrak

AK | In your upcoming project “Platform(n)ation,” you’re touching upon the current hegemony of optimization and efficiency at the heart of the rhetoric that surrounds the smart city. Metaphors such as “the city as a computer” can reduce our understanding of the city to a limited set of operations, measurements. Do you think there could be a better metaphor for a city?

NB | Yes, I think that quote comes from a piece by Shannon Mattern where she also acknowledges that the metaphor is misleading. It’s obviously a cliché, yet so many conversations about smart cities are focused on optimizing and improving such and such. But optimizing for what? I don’t think anyone who works with cities would claim that you can solve the city, or think about it in such terms. And even if we are to focus on the specifics of traffic flows or communication systems or pipes rather than “the city”, I’m still personally wary of thinking about it in this boastful and deterministic way. Optimization always happens at the expense of something else, with the premise that everything itself remains unchanged.

In terms of metaphors for the city, I definitely don't think the “city as computer” or “city as platform” are good or productive metaphors. Conceptually, I like to think of the city as the first five metres below the ground, plus everything that takes place above it. You can say that ultimately, it is organized service infrastructure that allows the city to function: everything above is enabled by these few subterranean meters where all the piping and wiring happens. There is also, of course, a layer of legal and social infrastructure, but I don’t think you can confidently claim to be able to optimize what’s above this underground layer. A colleague here at Strelka recently made the argument that, rather than always reverting to the image of the “smart city”, we should be thinking of “thoughtful infrastructure”, perhaps a more useful and actionable analogy. Rather than optimizing for x, we should focus on making our infrastructure more agile and adaptable—both technically and socially. You can’t fix a city, but you can make sure that its supporting infrastructure can evolve and respond to expected and unexpected change at the necessary pace.

AK | Today's cities are increasingly governed by the logic of preemption, pre-calculation, crisis management. Is there a metaphor for that?

NB | This is really a key design question to which I don’t think we have satisfactory answers or metaphors for yet. One entry point perhaps is hinged on the conversation about Google’s/Alphabet’s foray in urban planning with the Quayside Toronto project. Beyond both the predictable and somewhat banal cheerleading or resistance to the project, something that really interests me is the Replica tool that Sidewalks is building simultaneously to the actual development. Along with the real borough, they have a simulation model of the borough, and as they are building this neighborhood they are also building a computer model to track how circulation flows of cars and people might work, anticipating and alleviating congestion, accidents, unexpected events. I don’t think they are thinking about it in these terms, but you can see the projection of the city in the form of Google Maps not as a static representation, but as a cinematic model: a repository archive of the past, an operational diagram of the present, and a potential simulation platform for the future. This understanding of site and mapping is actually quite radical, and something we anticipated in GoogleUrbanism in 2016.

In the case of Replica and Quayside, Google is of course adamant about the anonymized nature of their model and the fact that they are absolutely not going to monetize their investment through advertising, but rather by building and eventually licensing better tools. But whether through advertising or other means, I think the value of such a system lies in its predictive capability. So the “city as a model”a simulation replica to preserve stability and order in the face of chaos and changing social, economic, environmental conditionsto my mind, is a useful metaphor, along with all the problematics that come with such a managerial vision of the city. Actually, the fact that it brings to light problematics that we can recognize is what makes it particularly useful. As every platform is developing its own totalizing model of the world, which includes some things and excludes others, we can address the idea of competing exclusivities and overlapping realities at the scale of the city.

City as a Model, 2018. 2018. Image by Anastasia Kubrak

City as a Model, 2018. 2018. Image by Anastasia Kubrak

AK | With GoogleUrbanism you proposed to introduce a new feedback loop between public spaces, users, and revenue. The idea you put forward is to link value, generated by the online activity of users, back to public spaces, where users are physically located. Today, cities are comprised of different real-time feedback loops: for example, Google’s Popular Times feature or the Uber surge pricing algorithm. But these loops often fail: for example, many Uber customers actually move away from zones that have a surge in price, which leads to a drop in demand, and drivers don’t even bother driving into surge zones anymore. A different kind of feedback loop emerges in such cases.

NB | Yes, I’ve heard about this phenomenon. Another example is Google Maps and other mapping platforms that may essentially be making traffic worse: everyone starts using the most efficient routes at the same time, which makes them more crowded. So, paradoxically, the circulation of cities potentially gets worse when we’re trying to optimize it.

Perhaps what we’re going to see is a personalization of these feedback loops that we were referring to earlier. Google Maps would propose individual routes to individual drivers—an uncomfortable proposition, obviously, that raises a lot of new questions. Some people are also afraid of these logics being applied in other ways, for example in a form of surge pricing and personalized costs for different consumers. If the supermarket knows that you are willing pay more, bananas might cost more for you. The highest price you’re willing to pay for bananas is what you, yes you, will pay for bananas. And perhaps that would somehow change your relationship to bananas? This is a tongue-in-cheek example, but I think it hints at the dynamic and open-ended emergence of feedback loops upon feedback loops, which I think is quite interesting.

AK | How do you see the position of the architect in relation to feedback loops? In your “Data-Nation” article, you address the role of the architect as a designer of relationships between things, and this could be an example of such thinking. Could feedback loops be deliberately designed to introduce an alternative logic?

NB | I think the reason why feedback loops at least appear to work, is that they must be simplified in order to follow the logic of the market. GoogleUrbanism is still a market-driven feedback loop. Generating advertisement to consume attention is still a market-driven exchange, and we tried to engage with this. What could be the externalities of this loop?

When designers create new feedback loops I think we become naïve: by trying to introduce and force alternative market logics, the loops become less loopy, more contrived and ineffective. It’s not to say that designers shouldn’t try to intervene, of course they should try to provide alternatives. But as with more traditional understandings of “site”, we shouldn’t think of loops as a tabula rasa. It’s sometimes more interesting and fruitful to use existing loops, because they work and can be critically examined. What are the potential side effects of these feedback loops? In the case of GoogleUrbanism, if people are already looking at their phones in the metro, what is an externality of that process? We can think of the metro system as a different kind of territory for design, which can be funded and maintained through a realignment with the feedback loop of monetized attention, as opposed to designing a new economic model from scratch. I’m not a purist about this. Currently, acting upon the reverberations of existing feedback loops is more interesting to me.

GoogleUrbanism, 2016. Image by GoogleUrbanism team

GoogleUrbanism, 2016. Image by GoogleUrbanism team

AK | Can space itself be a user?

NB | The notion of space as a user goes back to the idea of the user as a composite, a bracket condition. It is defined in relation to its addressability by a platform or system. It’s just a different way of saying that everything happening at a certain point in time, in space, is grouped together under the umbrella of the “space”. I think this is a useful shift away from the anthropomorphic bias of “user-centered design”, as the city involves not only people, but also systemic processes that are difficult to engage with at the arbitrary scale of the individual. It also allows for the inclusion of other forms of value in the form of data and unaccounted revenue from non-human actors, such as driverless cars, mobile billboards, or I.o.T. devices that exist in space. Aggregating and spatializing the user may be a way of addressing the ongoing thorny issue of data privacy and anonymity, as well as returning some of the unclaimed value that is generated by private tech companies.

AK | So space becomes a wallet? Could a different logic be introduced here as well?

NB | Well, one reason we focused on Google is that it operates the most used and comprehensive mapping platform: Google Maps. But if it weren’t Google, if it were an alternative system, that system would also stem from both a spatially and temporally contextual map of the city. That's what Google is good at: they know what to show, when and where. The spatial element is implicit, and the temporal element is at the forefront. I think this is an important shift in thinking about space and site as a moment in time.

I also think that as architects, we’re used to working with a map or a plan. But the definitions of both are changing. We no longer should be thinking of buildings as objects in space and then be done with it; there’s a possibility (and perhaps a responsibility) to follow buildings in time, know who is using them and how, keep track of their waste, be accountable for their repairs and externalities over time. The added temporal dimension shifts us from spatial design to contextual design in very pragmatic terms. So space as a wallet is another way of thinking of space as a context: one that is active, temporal, dynamic, and responsible/entitled to its own value.

AK | When we talk about strategies from an architectural perspective, there seem to be three stakeholders in this process: architect, state, and Google. How do you see the role of users in this equation? Can users enter this conversation or negotiation, or are they destined to wait for solutions proposed by architects and designers?

NB | As said before, I see the users as designers or co-designers through participation. If you design a protocol that no one uses, at the end of the day you didn't do much. So the users are stakeholders both through their activity and their feedback. If specific users launch initiatives, take part in the process and so forth, I absolutely see them as designers, even though they might not be traditionally trained in this way. But again, and especially because this isn’t the popular or prevalent position at most design conferences, I think it’s incredibly important to maintain a balanced and unromanticized view of bottom-up initiatives. There are many examples of users and citizens assembling and creating great things, but we should be aware that this process is one of hard work and far from being an automatic and smooth default condition. Fundamentally, I think the empowering quality of these platforms is that they allow the user to fuck shit up, and I mean this both in the good and bad way. Toppling down regimes and disrupting the status quo in decentralized ways is only half of the equation. The creative power of network effects is perhaps enacted in different ways.

How can users be empowered? If these protocols had a layer that was flexible enough to accommodate change quickly and take in different viewpoints, that would be the best-case scenario. Then again, if we are talking about integrating every possibility and perspective at the level of the protocol, it becomes too generic to be binding and useful. I don’t have a solution for this: I guess it’s a case-by-case consideration.

GoogleUrbanism, 2016. Image by Don Toromanoff

GoogleUrbanism, 2016. Image by Don Toromanoff

AK | How can users become self-aware and realize their positions outside of the citizen/state paradigm? In GoogleUrbanism, you deliberately use mainstream advertising language that could be easily criticized. But it also communicates to a much wider audience.

NB | In the context of my projects, I like to use vague and accessible communication strategies as a challenge and design parameter. It’s hard to simplify ideas and sometimes there’s a certain academic pressure to making them pure because we always want to appeal to our peers, to use that specific jargon and language, to have internal consistency and so on. But there’s a certain power in half-formed ideas. They are the ones that stir the political imagination and catch on more potently than fully developed ideas, which are difficult to access. And your own idea will (or should) change over time, so it’s more interesting to have one that’s not fully consistent, but rather malleable and shareable, so that people can add to it in dialogue.

The beauty and also the danger of half-formed ideas is that even if they’re not fully functional or true, they still work. If people believe that they are true, or make them true in their own ways. If citizens/users are to be empowered, they need access to ideas that are not fully formed, with enough internal contradiction to be stretched and accommodated by different users. For example, with our project we wanted to hammer down the idea that you’re in the physical world before you’re in a virtual world, and that platforms that use and leverage space should give something back. It’s not a very scientific argument. People argued, disagreed, and understood it in different ways. But it allowed users to acknowledge their position in this process in a different way, beyond the traditional frames of reference. We tried to create a correlation between users, platforms, and space that invites a reimagining (or at least a re-questioning) of the citizen/state paradigm.

AK | So design projects should go beyond utopian or dystopian thinking?

NB | I think working with projects that are ambiguous enough to accommodate both dystopian and utopian notions is not only more interesting, but also much more useful from a design perspective. In the real world you cannot impose righteous utopian principles, and that imposition itself is a very dystopian prospect. What I personally find most frustrating about some critical and artistic communities is the obsession with being “in the right”. When that becomes the end goal, we are forced into the vicious cycle of critique upon critique, which is an antagonistic and ultimately demonstrative mode of practice that sometimes needs its audience more than the other way around. The same goes for the intended proliferation of dystopian scenarios as warning mechanisms: what is meant as food for thought ends up becoming a shared image of the future that we go about reenacting. This is why I would say that a fiction is not necessarily a final design output. A design project is not about narrating future fiction a or future fiction b. Both futures are contingent, and both futures are implicit in every design project. In my view, and also in our work at the Strelka Institute, the more you look at a project and the less certain you are about whether it’s a utopian or dystopian proposition, a great or truly terrible idea, the more powerful the project is.

AK | Fiction is not a design project. This is an interesting thought, especially in the context of art institutions like the Sandberg Instituut, where definitions of design are pretty blurry or liquefied.

NB | I mean that fiction is perhaps more of a tool than a design outcome in itself, an ingredient in the mix. Arguments and “common sense” are always products of fiction, and even facts are always “faction”. What I mean is that as a designer, I’m not sure my role is limited to creating fictions about the world, either as warning signs or as advice. It’s a methodology used in order to allow and anticipate different ways of getting into a project. I think I also may be sensitive to the fact that many design departments seem happy to be transitioning into art and activism departments, creating a vacuum at a critical time when actual design (not “design thinking”) is very important. Advice and warning signs are also important, but advice and warning for whom? Presumably for designers, and that’s us.


The interview is an excerpt from Anastasia Kubrak’s MA thesis project User-Agent: If everything is so smooth, why am I so sad?,” which investigates what it means to be a human User in today’s technological infrastructures.


Nicolay Boyadjiev is an architect and design strategist based between Montréal and Moscow. He is working at the intersection of architecture, infrastructure systems, and platform urbanism, with his recent projects focusing on addressability and cognitive extraction at the urban scale. His design and conceptual work has received multiple awards and has been showcased in Volume, Moinopolis, Fast Company, and Architect Magazine. Nicolay is currently the Design tutor of The New Normal program at the Strelka Institute for Media, Architecture and Design in Moscow.

Anastasia Kubrak is a designer and researcher based in Rotterdam. In her work she focuses on political implications of emerging technologies, surveillance economies, and algorithmic governance in urban space. Her work has been exhibited by VanAbbemuseum, Baltan Laboratories, WORM, Chamount Biennale de Design Graphique, and her writing has been published by The Institute of Network Cultures and Nichons-nous das l’Internet.