“A symbiotic interaction between humans and machines”

Q&A with Dr. Frauke Zeller

Dr. Frauke Zeller (Ryerson University) believes that the Vessel could change the perceptions we have about robots, for example that robots cannot have a mind of their own. She is particularly interested in finding out how people are going to respond to an artificial being designed to be aware of its own unique identity.

You are a professor with Ryerson’s school of professional communication and have been studying human-robot interaction through the lenses of communication and linguistics for many years. Why are you interested in this particular project?

This project is particularly fascinating because it really starts to question the concepts we used to have about robots, for instance that a robot can’t have consciousness. This makes it challenging, at first, because all the traditional categories we have been applying in our methods don’t quite work anymore. So I am interested in defining new ways of looking into social robots.

I mainly work in the social sciences and have conducted many traditional human-robot interaction experiments. That’s why I also know how the STEM disciplines [the disciplines of science, technology, engineering and mathematics] approach these areas. And I am always interested in bridging these two fields, the STEM with the social sciences and the humanities, and see how they can come together, especially in this project.

You are a professor with Ryerson’s school of professional communication and have been studying human-robot interaction through the lenses of communication and linguistics for many years. Why are you interested in this particular project?

This project is particularly fascinating because it really starts to question the concepts we used to have about robots, for instance that a robot can’t have consciousness. This makes it challenging, at first, because all the traditional categories we have been applying in our methods don’t quite work anymore. So I am interested in defining new ways of looking into social robots.

I mainly work in the social sciences and have conducted many traditional human-robot interaction experiments. That’s why I also know how the STEM disciplines [the disciplines of science, technology, engineering and mathematics] approach these areas. And I am always interested in bridging these two fields, the STEM with the social sciences and the humanities, and see how they can come together, especially in this project.

What are you hoping to learn from this?

I think it will be very interesting to find out how people are going to respond to a robot that has some features of artificial intelligence. How will the Vessel change our perceptions and understandings about what a robot is? Will there be some kind of different symbiotic interaction between a human consciousness and an artificial consciousness? And how will artificial intelligence influence human behavior?

 We are also learning a lot from each other and our different research contexts since our project is not only interdisciplinary but also an international collaboration – and I find that very exciting as well.

Experts and scholars from various fields are participating in this project, not just AI engineers. Why do you think it is important to involve experts from different disciplines?​

What I really love about this project is that the core team comes from the arts, humanities and the social sciences. So right from the start, there has been this creative team looking into designing a robotic experience. Of course, we will have a lot of people from the STEM sciences working with us and advising us, and I am not saying STEM people aren’t creative. But our approach has been very broad in terms of what we want people to experience at the end of the day.

In the STEM disciplines, what we often try to do is to cut everything down to one specific small example that we then put into code. For example, if one feature of consciousness means being able to see, one way to simulate that is looking into how the robot can recognize the people it is interacting with. Whereas we, coming from the humanities and social sciences, are trying to have a very broad approach which of course makes it more difficult. We are always trying to have the overall idea in mind, which is the interaction of humans and robots and how they can tell stories together.

I love this idea of generative storytelling, that whole notion that Page [Richards] is bringing in. It will be interesting to see if and how humans and robots can engage in mutual storytelling. So far, we usually try to have this very strong distinction between master and slave when it comes to human-robot interaction: the master being the human and always being in charge and making sure that the robot or AI isn’t taking over. But how about having them as peers, and seeing what kind of stories can be told? That will be interesting to find out.

Some experts explicitly warn about the dangers associated with AI. Late physicist Stephen Hawking said the emergence of artificial intelligence could be the “worst event in the history of our civilization.” Do we need to be afraid of the Vessel?

No, we don’t need to be afraid of the Vessel. But we should never stop asking the question. And yes, it is true that we have recently made remarkable advancements in terms of that machines can learn from each other without humans interfering anymore. But this kind of AI that resembles human intelligence, that is something that’s not quite there yet. And I think that will take much longer than we think. For example, we are still struggling to teach machines the ability to understand the richness of natural language, like irony and humor. What is important, even if AI might outsmart us at some point, is that the agency always lies with the human beings that started the whole thing. Also, coming from the social science and humanities, we are very concerned about the ethics of AI and robots. We would never want to create a machine that purposely fools or hurts humans, for example, so we certainly are following the traditional Laws of Robotics by Asimov, also in terms of data privacy. So yes, we want to create some very high-end applications. But it won’t be applications where at one point the machine is going to do whatever, meaning we’re not in command of it anymore. Should we come close to that, we will definitely reconsider what we’re doing.

You were also one of the creators of hitchBOT who made international headlines a few years ago when hitchhiking across Canada and other countries. What did you learn from hitchBOT?

I think we learned that people, in general, have a very positive approach towards robots and that they take good care of them. People appreciated hitchBOT because it was approachable. And they were not told what to do with it which also created a lot of creativity. Now, that being said, that experiment was already more than five years ago. Our perceptions of robots have changed. People today often think that anything that is “smart”, for example “Alexa” or “Google Home”, is a robot. Coming from the classic tradition of robotics, I would not call these objects a robot. So we also might find differences, at some point, in how people approach an old fashioned kind of robot.

And what do you think will be different with the Vessel project?

HitchBOT we set out, not knowing where it would end up. The Vessel, on the other hand, will be staged to a certain degree, so less geographically dispersed. The whole notion of meditation and storytelling will be new of course, too. Storytelling was part of hitchBOT in a way as well but on a completely different level.

 What’s also different is the idea of trying to create some form of artificial consciousness. HitchBOT, through its speech abilities, at some stages created the impression of having its own mind. I remember that whenever we left it on the side of the road, I would always pat its head as if I was saying goodbye. I did the same when hitchBOT was about to start its trip through the US. That’s when it said to me, “I think I changed my mind.” And that came out of nowhere. In retrospect, I should have listened to hitchBOT… We experience that a lot: Machines that appear to have something like a mind, which is basically all based on imitation. With the Vessel, however, we want to create some form of an artificial consciousness that we then can do experiments with.

You were also one of the creators of hitchBOT who made international headlines a few years ago when hitchhiking across Canada and other countries. What did you learn from hitchBOT?

I think we learned that people, in general, have a very positive approach towards robots and that they take good care of them. People appreciated hitchBOT because it was approachable. And they were not told what to do with it which also created a lot of creativity. Now, that being said, that experiment was already more than five years ago. Our perceptions of robots have changed. People today often think that anything that is “smart”, for example “Alexa” or “Google Home”, is a robot. Coming from the classic tradition of robotics, I would not call these objects a robot. So we also might find differences, at some point, in how people approach an old fashioned kind of robot.

And what do you think will be different with the Vessel project?

HitchBOT we set out, not knowing where it would end up. The Vessel, on the other hand, will be staged to a certain degree, so less geographically dispersed. The whole notion of meditation and storytelling will be new of course, too. Storytelling was part of hitchBOT in a way as well but on a completely different level.

 What’s also different is the idea of trying to create some form of artificial consciousness. HitchBOT, through its speech abilities, at some stages created the impression of having its own mind. I remember that whenever we left it on the side of the road, I would always pat its head as if I was saying goodbye. I did the same when hitchBOT was about to start its trip through the US. That’s when it said to me, “I think I changed my mind.” And that came out of nowhere. In retrospect, I should have listened to hitchBOT… We experience that a lot: Machines that appear to have something like a mind, which is basically all based on imitation. With the Vessel, however, we want to create some form of an artificial consciousness that we then can do experiments with.