A discussion with Heather Wiltse about things that aren't things
How understanding what 'fluid assemblages' are can help us design and create in better ways
You’re reading DisAssemble, a biweeklyish philosophy of tech newsletter aimed at those interested in creating better digital products.
A few times in this newsletter, I’ve brought up a term - ‘fluid assemblages’. Sadly, this fascinating concept is not my own, but rather Heather Wiltse and Johan Redström’s, who coined it in their fantastic book Changing Things: The Future of Objects in a Digital World.
The book captures how digital ‘things’ are more like fluid assemblages, radically different than existing, more tangible things:
‘It is nearly impossible to assess and understand what these things do based on physical appearance alone. Unlike our old stable and predictable physical things, these digital things are networked, dynamic, and contextually configured.’
Heather, currently associate professor in design in Umeå University, was kind enough to sit down and chat with me about fluid assemblages and how their existence has implications for how we design and build digital products.
Can you speak a little bit, in your own words, about fluid assemblages and what exactly they are?
So this is a concept to try to mark the difference in what we're seeing with the things around us. We're not using record players, we’re using Spotify, which is this completely different sort of animal. Things are becoming more fluid.
We have software updates and licence agreements actually telling us that features may be added or removed, even ones that we rely upon. Products are being constantly made, with a progressive optimisation.
They're becoming more fluid, and they're becoming assemblages in that they're not just things that we can really hold in our hands. ‘Fluid’ is really being superfluous, it was really just a way to try to emphasise and call attention to the fact that things are changing. But in addition to the fluidity, they’re also these assemblages, they're made out of a variety of physical and digital network components.
What about fluid assemblages interests you? Why do you think they are important to understand?
Originally, it goes back to me being a PhD student, reading all of this philosophy of technology kind of stuff. Seeing the examples that these philosophers were using was very interesting. Heidegger's hammer is, of course, the classic, or Ihde’s telescope, or the bicycle and social construction - all of these fantastic examples. But computing is so different from these kinds of things.
What is the role of these new types of things in our lives? How are they mediating the way we come to the world, and the world comes to us?
I think there are significant implications for design. I'm in a design school. So this is very relevant. The way that things are produced and how that changed with industrialization - this called the entire field of industrial design into being, and now we're now seeing a change in the way things are being produced, again, which is calling for another kind of design.
Also, if we're not properly understanding what these things are, then we can't properly understand what they're doing and the role that they have in our lives. There’s also an important political dimension, in terms of what it means for us, as humans, and how we interact.
In Changing Things, you discuss understanding fluid assemblages through how they come together in use. Is this the main way that we can understand and perceive fluid assemblages - via manifestation in usage?
I think it's also everything around that - what actually enables them to come together. Looking at what's designed is not looking at a finished product, but as a set of rules for producing something.
Or, we can take the lens of Heidegger about things ‘withdrawing into experience’ when they are used. That's become a common design goal, but also a really big problem, because so many of the things that we use are also fairly blatantly using us to produce data about what we do. And they're designed to just ‘disappear’!
For example, Google Home devices are marketed as just disappearing into your everyday life. But wait a minute, like, that's maybe not a good thing - we actually need some breakdown, we need to disrupt that ready-to-hand and actually make these things present in a way that they aren’t.
Absolutely. In UX design, there's this idea of frictionless experiences. Within this, there’s the idea that the experience shouldn’t be reflected on - just experienced. This is obviously very problematic, when the implications of what you're designing have important secondary and tertiary implications, or are change who you are as a person. However, every time you create a ‘breakdown’ to help the user reflect during their experiences, you interfere with a user’s goals. How do you think we can navigate this?
I think that's something that needs work, honestly.
The whole idea of incentive structure is very important here, because in a lot of ways, there's no incentive, unless there's regulation to do it. We've seen that happen with GDPR, where it is typically experienced as just an annoyance. What GDPR causes is not really ‘revealing’ as such, unless you actually go through and read everything - like a Terms of Service or what cookies you are opting in to.
So when designing something, we have to ask, ‘what do we want to make visible’? Perhaps it’s segmenting things out in time, telling the user: ‘now, when you set it up, you really have to take some time, rather than just being sucked straight into it’.
I think we somehow need to reframe how we understand these things to be not just things for us to use, but things that are mediating a lot of different relations and different forms of value generation.
You mentioned value generation. Can you expand upon that a little bit?
If we look at the typical sort of value configuration, if we look at traditional industrial design, there's a ‘thing’ produced that has value for people. Yes, all sorts of things caught up in that value. But there's some specific reason a person will want to buy the thing. And then they move on. It’s fairly straightforward.
But now we have these multi-sided arrangements. If we look at any time we use a social media platform or a shopping platform, we get value from it. So there's still a definite reason for us to use it.
But then there's also value generated for other actors in the system, including advertisers. And the platform, having that position of being the mediator and the actual grounds for these transactions to take place, can see what's going on. There’s all these potential constellations around the platform of other parties.
Instead of just this producer-consumer logic we have an ongoing relationship as well, between a producer and a consumer. You can't just buy Spotify, you can subscribe for a month, you have a relationship, they get to know you, they get to guess what kind of a mood you're in, and then sell that moody moment to an advertiser.
I find that it’s incredibly hard to ‘point’ to these types of digital objects - which is something you also reference in your book. I’ve written about how we can get to grips with them. As a designer, I always try to map these objects out, as part of user journeys, service blueprints, concept maps or the like. How do you feel about this ‘mapping out’ of digital products?
It’s something I’ve been thinking about for a while. How to do this is a huge challenge.
One vital aspect that isn’t touched on a lot is ‘how was the thing made?’ This beautiful visualisation of the anatomy of AI by Kate Crawford & Vladan Joler is a great representation of things like Alexa. It’s an in-depth look at all sorts of different aspects of what it takes to actually make AI.
But in terms of mapping out how products work, you run up against the issue of increasingly working with machine learning and artificial intelligence , where you just really hit a limit of what you can even follow. You can see an input and an output, but you don't know exactly what happens in the middle. So yes it's important, but it's incredibly complicated, I think is the short answer.
A key challenge that I find with ‘mapping out’ or visualising systems or experiences is that the map is not the territory. Sometimes I think we treat the map as more real than the territory. Or we create metaphors rather than maps, and these metaphors end up being ‘the thing itself’. The metaphors and map you use, and the implications of the metaphorical structure, can have implications.
Computationalism is one way to phrase it. But I think we have to try to get a model that is good enough to get you something useful. In doing this though, the problem arises of how we as humans are modelled. We're data objects in these systems. We're defined by whatever categories and tags are used and whatever connections we're able to make and whatever reactions we can give to a post. That's how we're defined in these ‘maps’.
It’s interesting hearing you speak about how we are modelled within ‘maps’ of fluid assemblages. Do you think that it’s possible that we extend or externalise ourselves into fluid assemblages? Andy Clark, Dave Chalmers and many others talk at length about this. And, if so, do you think that this changes our agency, our cognition, or our identity?
Yes, this definitely a useful way to look at it - we're sort of externalising ourselves into these devices that help us get through the day. I use my calendar to remember that we were meeting and to get this zoom link. But then the other perspective is that it's not us doing and owning it. It's the platform companies. And they're the ones that are really incentivized, again, to try to figure out as much as possible about our everyday lives in order to predict who and what we are, and what we will do. That's the driver that I'm worried about.
In UX and design, we often talk about ‘vision’ and ‘outcomes’ over ‘outputs’. Is this helpful in terms of understanding fluid assemblages, and perhaps mitigating these more negative aspects of fluid assemblages that you are discussing? I’m thinking of how this can help identify the effects of something rather than just focussing on the thing itself.
Whether you define it or not, a vision or outcome is going to be there, it's just going to be implicit or something that you work with intentionally.
Ultimately it goes back to the incentive structures and the economic models that we have, that define those sort of end goals the or the metrics that you're going to be checking for. Very broadly, they are going to be around engagement, because it's that constant engagement that allows for fuelling the production with behavioural data. If you're working in that kind of a context, you're always going to have that basic sort of frame within the current kind of a system.
To me, it’s vital to frame a vision as both a proposition and question. This is what design is. I think it’s important to hold on to the intention, hold on to the implementation, but a little bit loosely, even as you put things out into the world that are concrete.
So what do you think that a designer's, a developer’s, anyone who works in tech’s responsibility is in terms of these amorphous, indistinct objects?
There's maybe a hard limit in terms of just your character or in terms of how far someone is willing to go. Otherwise, I think there is a responsibility for them to just understand what they're working with. One concrete example of this has been the discussion around dark patterns and interaction design, using design skills to further a kind of manipulation of users. I think there is a responsibility to avoid that kind of practice. I think these people have to understand what's going on, and at least, lift these questions up as points of discussion and maybe propose alternatives.
If designers can do that, like really understand what they're working with - and this goes back to the ontological and conceptual - you understand what you're working with, then you can be much more sharp about it.
And this becomes all the more important as fluid assemblages get ever more amorphous, with things like blockchain and artificial intelligence. Where do you think fluid assemblages are headed in the future?
It’s clear we have forms of AI that have agency - so this is a vital area for design to think about. How can we work with artificial agency? How can we work with assemblages that are doing things with us?
In terms of blockchain we have to ask what we want. I haven’t heard much that sounds amazing. Two things make me not enthusiastic. The first is that it’s an architecture that doesn’t require trust. This is depressing to me, as I think we should be moving in the opposite direction, especially given where the planet is. We need to be figuring out how to build trust and be vulnerable to each other. The second is the huge environmental implications from these huge calculations. Again, given the situation we’re in, this is a dealbreaker for me. Someone could convince me otherwise, but this really doesn’t seem like the thing we should be investing in.
Thanks Heather for this elucidating chat. Be sure to follow Heather on Twitter!
Heather is currently associate professor in design with a focus on the data-intensive society at Umeå University. Her interdisciplinary research centers around trying to understand and critique the role of (digital) things in experience and society in ways that can inform design, and it sits at the intersection of design studies, philosophy of technology, and critical technology studies.
She has published and/or presented refereed work in philosophy of technology, science and technology studies, human-computer interaction (HCI), and design research. She is part of the DCODE project team and a member of the executive boards of the Society for Philosophy and Technology and the Design Research Society. Her most recent book is Relating to Things: Design, Technology and the Artificial.