10 - predictive systems shaping an ownership economy
As an update on the activities of Cities of Things, I am very happy to mention we got the news to be selected for a CityLab010 project. Find more information on the website. I use one of the cases we will use in that project in the post below as an example.
In this reflective blog I like to dive into one of the fundamental concepts of Cities of Things that is touched upon in several posts but deserve a specific fleshing out I think; the active and initiating role of the bottom-up based network of objects that builds a Cities of Things, what makes a Cities of Things stand out other smart city concepts. It has a lot of aspects that can be dealt with in several posts. Like the role of relations as defining element, and the connection to incentified systems.
Back in 2012 when we kicked off the INFO innovation lab, big data as an important new concept was a rising topic. I reflect on the developments in a whitepaper and promoted an angle where big data served the creation of tiny services, meaning that you should not use the big data as just profiling, generalizing, analyzing tool, but keep it connected to the personal interest of users. I defined three separate layers to support that focus: the objects and the relations between the objects, the people, and the relations between the persons, and binding these together in an intelligent orchestrating layer that generates bespoke services via rules.
I had to think about this while formulating one of the core concepts in Cities of Things; meshed networks of objects that unlock a kind of intelligence in the dialogue of different actors, both human and non-human. The intelligence is not only meant for understanding and responding to situations but also for taking a lead, initiative, really autonomous behavior. That these networks of nodes are not only meant to cater to human interactions or relations but also have a responsibility to cater to the non-human actors, our fellow citizens, is an important notion. The presentation of Anab Jain and Jon Albarn for Dezeen last month is a great insight into that different way of approaching non-human or beyond human systems:
Move from systems to assemblages, from knots to nodes
Acknowledging the entanglements without the desire to have the “full overview”, keeps us open to surprising possibilities. And it reflects the deeply entangled co-evolution of humans and non-humans – think wolves, men and dogs, or the soil as a living organism.
In the field lab projects we are shaping within the Cities of Things Knowledge Hub, we are also using this new relation between human and non-human actors as a point of departure. Let's use the plans for the plan in Rotterdam as a starting point.
Here we take the community-managed waste system of Afrikaanderwijk as a case. What can change in a neighborhood if the waste system is not only a tool for collecting but also an active partner in orchestrating the collection process and taking the waste also as a resource by making the right combinations at the right moment. The system could have more automated collection devices on a neighborhood level and the system could use the knowledge of households’ waste status to make more sense in the waste circulation. Where you don't create a surveilled system where all trash cans in households get sensors; that would create a system that is intruding in the personal sphere.
What can happen though is a system that knows what it needs and will advertise to all connected peers that is it is needed for something. The peers can offer in a dialog the waste and gain even 'tokens' for this that can be used in other ways in the neighborhood.
The intelligence is more than simple threshold-based needs, it will use multiple inputs from the system, both in what is needed for the most efficient circular manner, and orchestrate the collection vehicles, etc. The goal should also be that the system with the users is learning by doing and creating a more efficient living.
What stands out here is the shift towards valuing the relations between objects and humans as key. This is the cornerstone of the model of predictive relations; as is sketched in the figure below. The constant interaction of humans and objects are creating a relation. That relation is what defines the value of the functioning of the object. With a digital component that mental model of the relation is influenced by digitized knowledge. From the past, from the programmed behavior captured in the digital twin of the object. And predictive knowledge that emerges from the relations happening in similar interactions at different times and places that are comparable and relatable to the interaction happening at the moment.
This is kind of abstract maybe so let’s try to carve out a concrete flow based on the mentioned case of a waste system in a neighborhood.
There is a constant relation between the human actor and the waste system. The human understands that when trashing something it is collected first in his own small waste container and at a certain moment it is time to put the waste container out for collection by the waste service. The way this works is captured in rules and expectations. The system behind the waste container at home has a mental model in the head of the inhabitant of the house.
Now imagine that the system is an active and intelligent actor. That knows about the waste levels in the bigger containers or even the ones at home. And knows when is the best moment to collect. Then the system might start a dialogue the moment that action is needed.
One step further in the evolution of the waste system, the waste will be seen as a resource for new production. And the collection of that resources is based on demand in the system. In that situation, the relation of the human and the system is changing from a mental model about the waste collectors responding to levels of waste, towards a role where the system is requesting for certain types of waste to bring outside.
Predictive knowledge can help to fuel intelligence and interactions. But you can think a step further; the system can predict what would be needed end of the week and suggest certain choices in buying produce to facilitate the circular loop. The essence of predictive relations is not about the predictions but about the disconnection of relations in real and virtual space.
Now we landed in a system of continuous exchange of values. In the dialogues, every side controls the exchange to a certain level. We are back to the beyond human responsibilities; the waste and the objects that will become waste have an equal vote.
This triggers of course all kinds of questions to research and design. In the General Seminar session of 20 December, we discussed Web3 systems, and as an example, we looked in the breakout to the refrigerator that did its own supplying. What is defining this? Is there a behavioral system that will learn from needs and adjust the supply? What if it gets disturbed with confusing impulses? Is then the only way out to influences to understand how to “reprogram” the system by designed behavior?
The framing of Web3 as the change from the attention economy to the ownership economy can be a useful angle too. In this article Li Jin and Katie Parrott state that four main ways that that happen, the most interesting for this case: by creating pathways for creators to own not just the content they produce, but the platforms themselves. This can be true for physical systems and things as the mentioned waste system. A variation of the ownership model that is baked in structures as DAO also stresses a participation model. A collective that feels responsible and is really working together, the value created for all parts of the collective that ‘owns’ the whole of the system… The building and tracing of the relations that in the end create the system are key.
Every time I discuss Cities of Things as a concept these elements of system design based on a form of exchange of values between humans and objects is the starting point. The value parameters are influenced or even defined by societal values. Or at least that would be preferable. There is a lot of work to do to make this understandable and applicable. A promising way to do this is creating prototypes, designing to learn.