“Psychologists can make humans and computers work better together”

Human-machine interaction has fascinated Chris Janssen all his life. Whether it's our behaviour in self-driving cars or the way people operate a microwave incorrectly, according to the psychology researcher, it says a lot about our future. He therefore thinks it is high time for more experimental psychology in Future of Work research.

Do we now, in innovation, use too little knowledge about psychology?

I think there is a lot of research into the effects of automation, for example, on groups of employees, but we also need to look at the individual. And that's where experimental psychology can make a contribution. How do people as individuals deal with a computer? And what do we learn from this about the cooperation between humans and systems? An important point, for example, is the so-called 'irony of automation': automatic systems that are developed to make situations safer, sometimes make a work situation, ironically enough, less safe. For example, if an automated system seems very secure, people will look less often at how the system works. Any errors will then be missed. This can make it unsafe.

Our attention wanes with far-reaching automation. I study how we can get that attention back.

What are other key misconceptions about human-machine teamwork that you want to dispel?

An important misunderstanding is, for example, what we call MABA MABA in our field. This stands for Men-Are-Better-At / Machines-Are-Better-At in task allocation. A classic mistake is that we think that computers are better at some tasks, such as fast math, and that humans are better at other tasks, such as creative processes. Those heuristics are then used to divide tasks between man and machine.

It has long been clear that it is not so black and white. There are computers that create works of art, that are constantly learning in a creative way. There is even an Artificial Intelligence song contest! With songs written by computers. In addition, there are people who perform logical processes better than computers. Dividing tasks between man and machine according to simple heuristics is therefore really too simple. Modern research tries to see man and machine as a team that complements and supports each other. So don't split tasks, but work on tasks together.

Working together as a team is also possible, for example, when driving a car. It is often thought that either the human or the computer determines the route. But it's better when they are a team. For example, a human can observe that there are unexpected road works and the computer can then calculate the best alternative route. If there are options, man chooses again.

Computers are becoming more and more creative: they make works of art. There is even an Artificial Intelligence song contest!

How do you collaborate with researchers from other fields?

Innovation often comes from bèta science. With new technology we can build something that didn't exist before: let's use it! But the builders of the technology are often not specialized in how a person would use this. Ideally, you will test it with humans extensively and long-term. Even more so, because the way we use technology is changing.

I think that touchscreens are a good example: when the first iPhones came on the market, we could never have foreseen that this technology would also be used in countless other places: for example on cash registers in shops, or on tablets in schools to teach children to write. It's hard to imagine a thing that is not yet created. This requires multidisciplinary research. Psychologists can help science researchers and product developers, for example, to test how people will react. And philosophers can help with ethical issues.

A very familiar question is: am I doing something wrong? Or is the machine doing something wrong? What do we know from research about trust issues?

Not nearly enough. There are many studies on how technology is used and whether it is trusted, but these are often studies over a short period of time; snapshots. Ideally, you also study the use of technology in one's work over a very long time to see the long-term effects. Because technology and its use, is a dynamic process. Both man and machine learn to interact with each other. A nice example is the self-propelled automatic lawn mower. It is programmed to make its round around the lawn every morning at 8:00 am to mow. It also has sensors that prevent it from driving or returning to its charging station if the grass is too wet. The owner sees that the lawnmower will not drive, even though it is 8 o'clock. He thinks it's broken and takes it back to the store.

For example, in the 1990s there were street interviews of people asking if they would like to use a cell phone: a lot of people said 'no'. If you were to repeat that research now, the answers would probably be different! On the other hand, it is the mobile phones of the past, not the smartphones of today. Calling is not the main function, not the main reason for people to use it.

Therefore, replication research is needed. What do you know about the different uses? Can we map that to see where things go wrong and where things go right with automation? Do we see the difference between use, misuse, disuse, and abuse? Everyday objects such as microwave oven still cause problems in use, so the design of user-friendly appliances is constantly evolving.

What will your research focus on in the near future?

Distribution of attention in the car remains one of my favorite subjects. Mobility is extremely exciting because actually all facets of psychology come together: making quick decisions, emotions, attention, etc. I look at which signals you receive as a driver, on the road, but also from the car itself. Is our reaction different, for example, if you use a light or a sound? And what sounds stand out when grabbing our attention? And what happens when you are distracted? I'm researching that.

We are also looking at other fields of work.Together with Stefan van der Stigchel, professor of cognitive psychology, I supervise a PhD student, Rutger Stuut, who is conducting research into the workplace of bridges and lock operators. How should you set it up so that it works more pleasantly and better for the operator?

I supervise a researcher who wants to improve the workplace of bridge and lock operators of Rijkswaterstaat: how do they view their screens?

In the past, the bridge and lock keeper often looked out the window to help ships. That still happens, but they now mainly look at screens with different camera images. Rutger mainly studies viewing behaviour. How is their attention distributed among so many screens? This research started in September 2021 and is part-time, because he is conducting it alongside a job as a consultant at Rijkswaterstaat. So the research is still very early, but very exciting. Especially that in this way we can work with real professionals from the field. If there's any news, I'll be happy to report it!

What motivates you in your work?

I like the melting pot of topics that come with human-machine teams: fast switching, cognition, decision making, attention, division of tasks. The way people and computers communicate with each other fascinates me and I want to make it better. For example, can the computer see when our attention is slacking and somehow get us "back on track"?

You want to encourage others to cone aboard?

Both in business and in science there is much more room to participate in research into automation. I'd say: Humanities and Social Sciences, let's hear from you! Make sure that you are valued and that your knowledge is included in innovation. It is not always easy to collaborate with a completely different discipline. Even the words you use: what do you understand by a “test”, “data” or a “model”? That can be very different in one field and in another. Yet it is important to build those bridges.