blog – co-performance of human and nonhuman

Due to my holidays, I write this post a bit later than usual. Welcome to the new subscribers! Every month I take one article in the domain of Cities of Things that trigger thinking about one aspect that is part of the Cities of Things elements of living together with intelligent things in cities that shape our city life so to say. That can be a new insight or a connection to one of the core aspects. This month I like to dive into one of the latter categories. One of the concepts I ran into while doing research at Delft University of Technology is the co-performance as described by Kuijer and Giaccardi in 2018. It reflects on a notion that we grow into a form of collaboration with technology that is based on shared goals, on a certain leveled interest almost. At least that is what I especially take out of it.

In the paper, the authors elaborated in detail. They use as one of the examples the way we change our relationship with heating systems over the years. From very functional and close to the material to a more distant technology. And now we got into an intelligent system with Nest as a typical example where technology becomes more of a partner in crime to regulate the heating. “Not only in the decisions made by the thermostat, but also in human judgments; today, numerical temperature values play a more prominent role in human judgment of comfort than in times of coal fires.”

What I like about the thinking of co-performance is the moving away from seeing technology as purely a tool for humans or an extension of the human capabilities, and introducing an agency level for the intelligent systems and things that are working as real partners based on shared defined goals.

In technology philosophy this is also a topic in a different way as a tool is analyzed as not only the tool to use by the human but also triggering the use in a certain way. Sometimes in a limiting way even, as we understand that every nail is triggering a hammer to just one form of use. I would not go into the depths of that now.

I had to think about co-performance while reading this article titled; Users Want Self-Driving Vehicles That Drive Like They Do, Sort of, based on research done in the US. To be precise, the article and research are not about co-performance, but it makes sense to look at the results with the co-performance lens and trigger new insights.

Quoting the article: “The results showed that most people prefer a self-driving car that drives like they do, only less aggressively. Participants who reported that they trust or somewhat trust artificial intelligence, autonomous technologies, and self-driving cars expected a car with behaviors similar to their personal driving behaviors.”

Next to Nest, I think that autonomous driving is a poster-child for our relation with intelligent things, it is the most far developed in maturity and intelligent objects are most part of our lives now. Next to vacuum robots but they are most of the time not so intelligent, smarter adapting.

Back to research; it is interesting how drivers apparently mirror their own driving style on the expectations of the driving style of the autonomous cars. As a better version of themselves. So in the sense of co-performance there is a kind of symbiosis expected between the own capabilities and that of the intelligent objects, which seems like a small step to shaping a co-performance.

It is a remarkable finding I think as you might expect upfront that we see autonomous cars as a separate thing that has an operating style on its own, optimized for the safest way of driving. Super solid and maybe boring functional, preventing any surprises. I think it could be much more interesting if the cars adapt to the human driver, and even more interesting if the operation grows into a shared form of operating, the better self, or maybe another form of driving.

It is interesting here to think about the interface between the car and humans. You see a lot of apps that are designed to connect to intelligent things making the behavior transparent and make settings controllable. Sometimes it is more a learning experience.

Two examples from some time ago that pops into my mind that I would give as inspiration to designers that want to create these symbiosis-supporting services. One is an intelligent coffee maker that was introduced on Kickstarter some years ago. It has all kinds of smart features, not to control the coffee maker from a distance (that was possible too) but mainly to create the optimal coffee profile. It offered insights into the working of the machine and possibilities to control these, combined with different choices in ingredients. The coffee machine lacked the agency in the process to consider it as a co-performance, but the interface and interaction dialogue would make it a small step.

Another example I used a lot to reflect on different forms of interacting is the haptic interface a student designed for self-driving cars. Instead of communicating via screens, the chosen interface was an element where you put your hand on while driving. It was a kind of lever, a controller that was in rest but could move translating the intentions of the car to the passenger. The passenger could intervene with the choices made by the car by moving the lever.

This feels very close to that symbiosis that is inspired by the concept of co-performance. It also links back to the research on driving characters. How would that translate to interactions you can do? Can you indicate your style or mood by the way you interact; from easy going to agitated. How will that influence both the trust in the car as the agency-level of the human-nonhuman system?

The article on the research was published in an insurance magazine. Without maybe directly meant like that, it makes a lot of sense to explore this. If the car is in co-performance with the machine who is in control? Bot machine and man. Is he always making the last decision? And if there is a legal problem, who is the one to address it? And can the regularly touching of the steering wheel in a self-driving car be replaced by more subtle interfaces as the haptic controller?

Enough to explore further. Check the paper on co-performance. In the meantime, I remain driving my thirty-year-old French car that does not have any of this technology of course but has on a different level his own will…