It was an honor to host Professor Hiroshi Ishiguro (Osaka University) and Professor Tatsuya Kawahara (Kyoto University) and their masterpiece Geminoid HI-5 as our research group’s guest speakers. Below is a special interview with Hiroshi Ishiguro, a principal investigator of the ISHIGURO Symbiotic Human-Robot Interaction Project, funded by ERATO.


Possibility of Making Android Movies

Matsuyama: You said you had a plan to make a movie?

Ishiguro: Well, if I would make a movie, it would be like “Ex Machina” that doesn’t cost much money. You see that we feel some kind of humanity in a robot when you deal with them. I’d like to express that kind of sense in a movie.

The most difficult part of human-robot interaction is that you understand it if you touch it or you talk with it, for example, you saw that our android moved yesterday’s workshop. It’s a whole different story when you see it for yourself, rather than hearing about it from someone or looking at pictures of it. I want to express that impact in a movie.

If it’s an ordinary movie of androids, there might be some storylines and we’d see a robot as an android, while our android head may be “too human” controlled by computers with a certain “existence (存在感 sonzai-kan)”. They don’t use real robots in movies, it shows halfway humanity under computer control and that makes a perfect presence of androidness. Well, it is not what I want – instead, I want to realize a robot-like presence that leaves awkward controls that can not be a perfect human being.

You know the “Uncanny Valley”, but if you do not have an uncanny valley you are in trouble. I was actually building androids thought that I had to avoid it a long time ago. But in a sense, I should have not avoided it. Without it, it cannot attract people, it cannot become unique. For example, my presence is somewhat uncanny from someone’s point of view, but you cannot become unique unless you are uncanny. In order for the robots and the androids to be unique, it is necessary to have some uncanniness in them. It would be ok to make them look completely human, but it becomes just a person who you cannot tell from one another. I don’t feel that can make a story, an android or a movie.

Geminoid HI-5 (Photo © Yoichi Matsuyama)

Uncanny Valley and Humor

Ishiguro: Recently, my perspective of the Uncanny Valley is changing, I am thinking about what it means to laugh, for example. It is not about cracking up but a humor which nobody is really studying as science. How much do people take humor seriously? Everybody laughs, right? Just as I talked about the appearance, everybody cares about how they look and it it’s quite an important thing, but nobody ever have studied it. So I started it myself. It goes same with the humor, it is nonsense that no one ever have studied it seriously. I think the humor is the principle of human evolution. Having a unique feature or humor means somewhat different from others, while nobody take it funny when that difference is quite big, but a human laughs at somewhat different things. In other words, it is a principle of evolution – we have to accept persons making us feel the sense of evolution. Such unique people might hold tremendous values. Androids should also be such things. Perhaps it may not be accepted in countries where people believe in the human supremacism, such as the United States or European countries, to welcome a new kind of species. In the meantime, Japanese people may not really care about that and accept them, just like welcoming a new friend, I guess.

“I think the humor is the principle of human evolution.” – Hiroshi Ishiguro

We are still making androids that look like real humans, but we don’t have to match details one by one between them, while it’s more important to make them being unique. It’s same with the humorous stories. I think we could see the real possibility of robots as new species when people see them unique beings, or robots talk about things from “the robot’s point of view” or something that cannot be said by humans. I wanted to make such a movie. I think it would be very interesting if I could make a movie which represents that robots may be the crucial beings contributing to human evolution with explanations based on the principle of human evolution, but not like many Hollywood movies where robots

Matsuyama: Why don’t you do it?

Ishiguro: We’ve already begun as a research based project. I am interested in making a movie, SF, or art because I know how much we can do within ten years. Having trivial research projects won’t move things forward. An advantage of making art or movies is that they can make one’s concept come true in reality quickly. For instance, we’ve had a lot of awards with the “Alter” project that we’ve been doing with Prof. Takashi Ikegami of Tokyo University recently [Ikegami et al. 2017].

In the fields of the art, you can’t get the first prize unless you follow their formats. The Android Theater (“Sayonara (Good-bye)”) exhibited in Ars Electronica we did got only the third prize, but it is the most popular one among people. People who lives in the fields of art, they don’t let any outsiders to destroy the format they’ve made. I’d like to do things which make us realize “what is life”.

Sayonara © Tatsuo Nambu / Aichi Triennale 2010

In the “Alter” project, we created a very complex CPG (Central Pattern Generator: a mechanism that generates periodic exercise such as walking and generating periodic exercise) and a humanoid neural network called the Izhikevich’s Spiking Neural Network [Izhikevich 2003]. In fact, we are not quite sure about the principle of the network yet, but when it’s executed, it becomes like a living thing. We took such an approach making our audience recall living beings that looks like they are really thinking. So it was a great success as art. Although it was the second prize of Ars Electronica, it’s an honor big enough for full-time artists to earn a living.

Matsuyama: By the way, are you conscious that you are somewhat uncanny yourself?

Ishiguro: No, I haven’t thought myself that way. But everyone says that kind of thing lately.

Matsuyama: You may have had this question frequently, but I think you have a style in a good way, such as your outfit. Do you have any ideas that you want people to look at you in a certain way?

Ishiguro: Do you mean my appearance?

Matsuyama: Yes.

Ishiguro: Well, I might be doing it without realizing it. When you ask yourself what the identity is, it’s same with a matter of appearance. I think the clothes are the one that shows one’s identity clearly. Out of three: clothes, a face, and a name – You don’t change your name, do you? You think it is your identity, right? The face may change from time to time, but it does not change that much. But you change your clothes often, don’t you? The clothes speak the most. When someone walks towards you from a distance, you see his/her clothes first, then recognize his/her face, and get to know his/her name only after you talk to him/her. So the identity by a name is not that strong in that sense. That’s why I haven’t changed the style of my clothes since I was a student.

Alter © Justine Emard

The Origin as an Artist

Matsuyama: What is the origin or formative experience for you as an artist?

Ishiguro: I always painted in a primary school, junior high school, and high school. I drew anything from landscapes to portraits, I only drew landscape paintings in the later period. I was awarded a lot of prizes when I was a child and I had some solo exhibitions until my third year in university. I attended some painting classes too. I had an idea to draw in the rest of my life if I would have not anything to do, but it was not that easy. At the time, the computers came out so I studied it eventually. Drawing a picture is like expressing humanness on the canvas, so I’m doing the same thing with the robots in that sense.

Matsuyama: What is your role model as an artist? Who’s your favorite artist?

Ishiguro: I liked painters like Corot, but who I admired was Chagall . I do not care about Picasso.

Matsuyama: Why do you like Chagall?

Ishiguro: Because I am the one who draws with reasoning, so that I admire the one who draws without logics. It’s marvelous that the one who draws only with the image came up on his head. If you look at Makoto Saito‘s paintings, you can tell the reason, but what’s amazing about him is that he draws without any education of the theory. I have been amazed by these people. So I was able to shift to computers without any regrets. The way I draw is like digging something up logically as much as I can, rather than draw with something popping up on my head.

Matsuyama: The idea totally makes sense to me. So you are creating robots as an artist.

I and the Village, Marc Chagall (1912). Royal Museums of Fine Arts of Belgium © RMFAB, Brussels

Ishiguro: Telenoid, for example [Ishiguro et al. 2011]. The sense that I value as an artist is intuition. It’s all the same after all, when I draw a picture, I draw with my intuition. But I’m the type who wants to explain the reasons of all the lines, why they are there. It’s same with my research – what’s important about the research is to create something by combining various elements first, then you have to explain the reasons afterwards. The image of Telenoids suddenly came up to me while I was sleeping at night.

Then, on the way trying to make that image come into reality, I asked model creators with telling “just like this”. However, none of them came closer to my image, they just couldn’t create the shape what I saw in my mind. After all, I decided to use a graphics software allowing us intuitively to model objects just like clay modeling. With an operator’s help, we managed to scrape out the Telenoid.

Telenoid © ATR + Hiroshi Ishiguro Laboratory

Matsuyama: So you didn’t make it by logically subtracting elements?

Ishiguro: No, I did not. That was not a subtraction. I was thinking about subtracting in the beginning, but that image suddenly came up to me. Reasonable explanations were added afterwards, for instance, you cannot tell its gender in a symmetrical design, you cannot tell its age with a proportion of a child body and an appearance of an adult face, while I had a concept of “neutral human” without any identity. Later, I made things like Hugvie that we intended to eliminate them thoroughly. So I don’t think anybody can make it.

In fact, there are many methods of subtracting such as stopping the movement of the eyes, letting them blink only halfway, that sort of things. We do them of course. For instance, ERICA is designed with some  subtraction methods [Ishiguro et al. 2016]. I’ve made a lot of robots trying to copy real humans before we adopted the subtraction methods. But ERICA is perfectly symmetrical, the more you subtract, the more it gets beautiful. So that ERICA’s face is gradually getting closer to Telenoid’s. The androids we are making lately are exactly based on subtractions. But, Telenoid came to me out of the blue.


Intuitions of Art and Science

Matsuyama: I truly understand what are you talking about. I originally intended to pursue career in art before. But at a certain point of time, I started studying robotics and computer science at a graduate school, while I systematically studied drawing at the same time.

Ishiguro: I did it quite well too.

Matsuyama: As you’ve mentioned, ones who can draw well can explain his/her current processes logically wherever he/she is asked. In fact, students from the Tokyo University of the Arts taught me drawing were explainable and they embraces the processes quite logically. I found activities of robot design and drawing have many commonalities.

Ishiguro: Yeah, drawing is logical when one studies properly like what to draw in what order.

Matsuyama: Principles of perspectives and rendering are also pretty logical. I don’t think it is art itself but a kind of technique.

Ishiguro: Well, the origin of art is technique in the first place. So I guess what makes difference is whether to use intuition. But it has to be explained logically. I think drawings of Makoto Saito or Chagall are pretty good and it should be explained logically in the near future. But what makes them special is that they create it before the explanations. In my case, we draw pictures with explanations along the way, so that we can explain the piece pretty much in details when it’s done. However, Chagall couldn’t explain it himself but everybody knows it’s great. I think the privilege that artists have is to find some new problems, but we scientists can explain these problems and let it have reproducibility. That’s the privilege of engineers and scientists. What’s wonderful about artists is that they don’t let their work to have reproducibility and don’t explain what it is, still people find their works amazing.

For instance, Professor Shinya Yamanaka was not able to solve the problem by finding the specific combination of protein unless he got intuition [Yamanaka et al. 2006]. I think everyone has that kind of intuition. You don’t have anything to depend on but intuition when you start doing something new. I assume people who don’t have intuition netiehr can be successful researcher nor artist. However, what’s severe about researcher’s job is that they have to let their work to have reproducibility with all the explanation. Otherwise, he won’t make it as a researcher. Why I stick to the robots is because the they are reproducible. In short, when one says  “I made something like a human being”, if it is a picture or an art, they would say “That’s something only you can draw”. But, if one makes a robot, they can copy it. Of course, you can copy the picture or drawings, but it doesn’t move and it is the only one. But robots autonomously do a variety of things that can be reproduced – they can secure their accountabilities only by showing their actions or other products – I think that is like a constructive science that can be regarded as a new science.

Matsuyama: I see. In that sense, what do you think about the exit of ERATO?

Ishiguro: ERATO is kind of a special program, in short it does not seem to have a goal. Do as you want, so they say. Rather, it is easier to work with CREST. It’s like “I say I would do this to that point, so I will do it to that point”. For example, we made Telenoid with CREST that was easy in a sense since there was a certain goal and we got a good reputation. In ERATO, an evaluator and a project leader are usually a same person. CREST projects can be evaluated by Research Supervisors, but ERATO projects cannot – You have to corner yourself to make something new. You can do whatever you want autonomously but that kind of freedom tends to lead the projects nowhere. But, this is something you cannot do with other fundings. For example, I thought that we should talk about the intention and desire quite frankly. Professor Kawahara places importance on the Total Turing Test and I think that’s a good thing to have it as a goal.

ERATO is the Exploratory Research for Advanced Technology research funding program founded in 1981 to promote basic research in science and technology. ERATO is one of the biggest research funds in Japan.
CREST is a funding program by the Japan Science and Technology Agency (JST) for team-oriented research with the aim of achieving the strategic goals set forth by the government. For each Research Area, a Research Supervisor acts as the head of a Virtual Institute.

Matsuyama: In that sense, you don’t think the Total Turing Test is important?

Ishiguro: Professor Kawahara thinks we should carry out the Total Turing Test technically as an indicator of performance. For example, how it functions in regards to  “attentive listening”. [Kawahara et al. 2017]. However, we, robot creators, see the Total Turing Test in a slightly different way such as how it gets closer to the human as a whole. Professor Kawahara takes the Total Turing Test as a tool to measure multimodal tasks one by one. From my point of view as a robot creator, we see the Total Turing Test in a broader sense. Our consensus to the Total Turing Test lies in the idea of how well it can communicate with humans.  Suppose we are going in that direction, the robot has to have his own intention and desire. Otherwise, it will be pointless. You see that the robots ever have made until now haven’t dealt with that matters. The robots don’t make decisions on his own but just do a given task or respond to what has being asked. I think the future of the robots lie only there.

If Amazon Echo or Google Home are not being used, the elderly probably wouldn’t use it as a tool either. I think it might be impossible for a dialog agent to exist, unless it becomes a partner or a companion, or it has a meaning of existence itself, instead of only a tool. Although the quality of Speech recognition and image recognition are much better than before, they cannot reach up to 100%. Are they able to buy things using Amazon Echo? No. Perhaps the “voice switches” probably will fail, I guess. If that kind of  technology would survive, it had to have a presence that is something meaningful just to talk to.

I also think it’s unnatural to place too much importance only on dialogue. It should be something like you enjoy talking while you do something together. It sounds more natural that it gets easier for people to do something with chatting with robots than doing it alone. Although we could cover elderly people only with dialogue capabilities, others need something in addition to dialogues.

Hiroshi Ishiguro (Photo © Yoichi Matsuyama)

Co-evolution of Humans and Robots – Conversation with Entities Making You Feel Evolution

Matsuyama: Do you have an insight on what motivation and desire should be put in robot as conversational partners?

Ishiguro: Well, it should be relationship building. People basically have social and personal desires. Ironically, while we want to be unique beings, too strong social desires may make all of us homogenous.

So why humans have interests in humans? It is because only entities or beings letting us feel evolutions interest us the most.  I think people would find it truly interesting when talking to robots when they feel “Oh, it might be an advanced creature ahead of me”. This kind of idea can be accomplished only in Japan at least for now. I don’t think the idea would be accepted in Europe or the U.S so easily.

Matsuyama: I found it’s pretty interesting. It might not be accepted in general though.

Ishiguro: Whom are we interested in? Whom do we want to talk to? The answer is that the ones who makes us feel the evolution of mankind, even for a little glance, right? So, if we can be interested in a dialogue with a robot, it must be that kind of being.

Matsuyama: It’s really an interesting idea. I really want to make some research ideas out of it.

Ishiguro: Don’t you think it’s important? But nobody takes it so seriously now. They may study “human robot interaction” and that’s it. Robots have to reach a level letting us feel the revolution of humankind. So it’s not “human robot interaction” any more, but it’s like human-robot “co-evolution”. We may be able to find robots really interesting only after they reach such a level.

Matsuyama: I’ve been also thinking about similar things like the meaning of the dialogue for a long time, and I came to a point where I found myself getting tired of making dialogue systems to be quite frankly. Of course, fundamental conversational functions, such as turn-taking, are important and we are working on to create systems. But, because I wanted to be a film director in the first place, I am interested in conversations as media. In other words, I am interested in conversations which are enjoyable on their own.

Ishiguro: What kind of dialogue systems have you been working on?

Matsuyama: I came to realize that conversations between people themselves are killer contents and I stopped making conversational models between a robot and a human since I thought it’s nonsense to replace a human with a robot that cannot live up to the level of the expectation with the current technology. So I realized situations where robots can augment existing conversations among people [Fujie et al. 2012] [Matsuyama et al. 2015].

Ishiguro: That would be like nodding or attentive-listening?

Matsuyama: Well, attentive-listening is closer. My major research topic was multiparty conversations. When I visited a day care center in Tokyo, I came to really understand that it is a fundamental need for elderly people to talk with others till their death. So I thought it was the right job for my robots to make their conversations more vivid.

Ishiguro: We’ve been working with senior people for a long time, too. But we have to stay away from the situations if you want to create something that change the world. Senior people are special cases – They cultivate their strong desires to talk and to connect with others while they don’t have much opportunities to talk since they don’t have sufficient staffs to assist. So they become like “children” in a way that they show their desire without hesitations. Those who  completely expose their desires actually can talk with robots just like children. Other ordinal people need attractive robots they irresistibly want to talk with, with hiding their desires. Conversations without showing one’s desires and attachments – We were just discussing this topic related to a new program of ERICA yesterday.

Matsuyama: Truly exciting. I’d like to bring home to think more about  “a conversational partner as an entity making you feel the co-evolution”.

Ishiguro: The robots won’t be recognized as something truly useful unless we think that far and deep. Pepper failed, Wakamaru failed before that, and Jibo failed as well. What’s the causes of the failure? In short, we take the robots’ presence with very narrow perspectives, just like seeing them as tools – that’s the reason. We should work on robots as if we are making new humans. Suppose we are asked “Why do we need such a thing looking like a human?”, we should answer “Because it’s the direction of our evolution.” I assume people would find that does make more sense.

Matsuyama: This is an unrelated topic, but I chose my wife with only one standard, “Is this person fun to talk with?”. I find myself the most fascinated when I talk with her. It is my research that I keep on thinking about why this is interesting. I believe each trivial dialogue we have is the very edge of the research.

Ishiguro: That’s a good thing that you enjoy being home.

Matsuyama: Exactly. For example, while we were waiting for a connecting flight at an airport before, we played a shiritori. (a word game in Japan in which one player has to say a word starting with the last syllable of the word given by the previous player.) It was a simple conversational game where we were exchanging single words with guessing what the other is thinking, and utilizing the timings of turns. I could not stop thinking why this was fascinating. I’ve been thinking how I can make such a converational interlocutor, however, I did not have a clear concept of evolution at that time. This idea really hit my imagination.

Ishiguro: How long have you been together?

Matsuyama: We’ve known each other for long years, but it’s been about 4 years since we got married.

Ishiguro: How do you feel? Don’t you find it boring?

Matsuyama: I never get tired of conversations with her.

Ishiguro: Is she pretty intelligent?

Matsuyama: Yes, I think we’re on the same Intellectual level in some sense.

Ishiguro: Like a coworker?

Matsuyama: Yes.

Ishiguro: Is she a researcher?

Matsuyama: No, she is an ordinary person.

Ishiguro: Did you study together in the same university?

Matsuyama: No, we didn’t. But we have the common friend who introduced her to me as “There’s a girl who is a bit different”. I think she is different in a good way. Because she is full of curiosity, I usually judge whether I go with a certain research idea by talking my ideas to her and see her reactions.

would take human’s work away and in the end every robot was destroyed and we’d be happy.

“Geminoid HI-5 is a tele-operated android head that has similar appearance of original person, Hiroshi Ishiguro. Geminoid HI-5 has sixteen degrees of freedom that allows itself to behave like actual humans. By using Geminoid HI-5, we demonstrate to questions that “What is a human presence?” and “Can human presence transfer to a remote place?” – Hiroshi Ishiguro Laboratories

Geminoid HI-5 (Photo © Yoichi Matsuyama)

The Darkness of the Mind

Ishiguro: There is always something that we cannot explain, isn’t? The one who touches something unexplainable is very important. I take it as the darkness of one’s mind. People who can explain all their hearts out to others are extremely social, and maybe you think they are reliable in a sense.

The human being is full of inconsistencies. Originally, social desire and personal desire never balance out. In order to evolve, one must be better and stronger than others to stand out. But if we only have that kind of people, the society would fail. That being said, one has to stand out while we keep the society functional. Our minds have contradictions in the first place.

And it is ethical, too. Some people might tell us to cooperate with the others while we all are fighting each other to survive? Why sociability matters in survival games? The way of survival of human species we found is contradicting from the individual point of view. But I guess it is not inconsistent from the gods’ point of view. Individual persons are fighting with contradiction while they don’t understand themselves.

To answer the question why sociability is installed in humans, my opinion is that it’s because we cannot model what’s inside us. We cannot get to observe what’s going on in our brains or in our stomachs. When we talk about our brains, we don’t really know if consciousness exists inside us, but we can tell they do by observing other people. There’s no way to understand one’s self without interacting with other people. But, at the same time, you have to outsmart them as a result of the interaction. That’s the contradiction and you are working hard to look into yourself.

If inner minds are quite simple, the evolution cannot occur. There has to be something, something you cannot really explain, otherwise, you’ll end up being just “one of them”. Having someone who can have dialogues with you, stimulating the darkness of the minds,  is a huge advantage. Quite a few people can have this kind of conversations.

Matsuyama: Very few?

Ishiguro: Yes, it’s very few and even if you have someone who has a good intuition, he/she may not dig into it deeper. I’m envious of you that you have a partner who can stand that depth of thinking.

Matsuyama: Well, I don’t really know if my partner is that good, but I do like to talk this kind of topics which cannot be a research or a paper right away.

Ishiguro: Probably, what I have said here is similar to biological principles, like a dilemma we are born with. Research won’t go anywhere without being rooted in such principles. Our question is whether we could go there.

Matsuyama: I see.

Ishiguro: If I made a movie, I think I could express that kind of thing.

Matsuyama: Sounds interesting.

Ishiguro: It may end up being a difficult movie like a French one that you do not understand at all. (LOL)

Matsuyama: Well, I think it’s still good to depict the existence of human beings with a contradiction as it is.

Ishiguro: I think that is what it is at the bottom of robots.

Matsuyama: Otherwise, it won’t be exciting.

Ishiguro: As I am talking like this, I can get my mind clearer, and get a strong sense that cluttered ideas could connect with the other. Now I think I’ve got a storyline of a presentation I’d talk.

Matsuyama: That’s good. I’d be happy to be wherever you need someone to talk to (LOL)

Ishiguro: In the recent lecture that I give to the general public, I talked about “humans will be robots in a thousand years” in “the last lecture” (a TV program of NHK-BS1). The content was much shortened in the TV program, but this is what I see about the future of robots.

Matsuyama: That’d be something what dialog systems haven’t met their true breakthrough, and I really think that’s what it is.

Ishiguro: When it breaks, it really breaks since we do a lot of collaborations with various companies. When one breaks, it breaks big and fast. And I think that it is the moment when the idea of “what is humans” changes drastically. The concept of countries has changed considerably due to the Internet. There is no country any more in that sense. Globalization has been spreaded drastically. When robots really emerge, the definition of human beings should change all at once. I think “individuals” will change. The concept of the country has changed because of the Internet, and individuals will change because of robots.

Research Changing the World

Ishiguro: I think it exactly is the evolution that the people with prosthetic limbs run fast in the Paralympic Games. There are some people who don’t want to get into that field, yet. Well, at least as a researcher, you have to do with that kind of mind. Some scientists might do their research only because it is popular at the moment, without asking fundamental questions about their purposes and goals.

I usually have these kinds of discussions with philosophers. I’ve been hiring some philosophers – it may not be unusual for a professor – at least one every year. They are truly helpful since their writing abilities are really excellent. They are helping me with writing quality grant applications. I usually tell them to do whatever they want to do as long as they could help us within a half year. All the philosophers I’ve hired have achieved successful careers later.

Matsuyama: That’s really interesting. By the way, this is also a personal matter, but I’m finishing my postdoctoral fellow period and am moving towards real independence to create my own field. Could you give me any suggestions that I have to keep in mind?

Ishiguro: Well, I haven’t thought about it that much in my case. But I can tell you two things that I’ve been told. Professor Saburo Tsuji told me to think about the fundamental problems all the time, which was his important teaching. Professor Toru Ishida from Kyoto University told me to do something that changes the world, but not work on just writing papers. These are the most important teachings I’ve got.

Matsuyama: What does it mean by changing the world in this context?

Ishiguro: As the Internet changed the world, let’s aim for something that changes the world. For whatever reason, we tend to be short-sighted. So it might be a good idea to do research, for example, with betting that the robots are the next human evolution. But if you advocate this idea vigorously, you might not be understood easily. Still, I believe you have to have that kind of belief at the root.

Matsuyama: It was really a stimulating conversation. Thank you so much for your precious time.

Hiroshi Ishiguro, Tatsuya Kawahara, and Yoichi Matsuyama (Photo © Yoichi Matsuyama)
Geminoid team and ArticuLab (Photo © Yoichi Matsuyama)

Project Details

ISHIGURO Symbiotic Human-Robot Interaction Project

  • Mission Statement: “Invention of Interaction Technology involving various communicative means as human use them, and Development of Autonomous Robots capable of naturally interacting with humans ranging from children to elderly in social contexts.”
  • Brief description of the project: “This research area engages in the development of Symbiotic Human-Robot Interaction, and its goal is to develop autonomous social robots that can communicate with multiple humans via various communicative means as humans use them. In order to achieve the goal, it is necessary to develop certain devices and technologies: (a) surface skin material and internal structure for safe interaction with humans, (b) robust and flexible speech recognition technology, (c) functions of autonomous context- and task-sensitive communication on the basis of a hierarchical model consisting of desire, intention, and behavior including speech acts, and (d) functions of using multiple communicative means to communicate with multiple persons in social contexts.”


  • [Ikegami et al. 2017] Itsuki Doi, Takashi Ikegami, Atsushi Masumori, Hiroki Kojima, Kohei Ogawa, and Hiroshi Ishiguro, A New Design Principle for An Autonomous Robot. In Proceedings of the European Conference on Artificial Life 14, vol. 14, pp. 490-466, MIT Press, 2017.
  • [Izhikevich 2003] Eugene M. Izhikevich, Simple Model of Spiking Neurons. IEEE Transactions on neural networks, 14(6):1569–1572, 2003.
  • [Ishiguro et al. 2011] Kohei Ogawa, Shuichi Nishio, Kensuke Koda, Giuseppe Balistreri, Tetsuya Watanabe, and Hiroshi Ishiguro. Exploring the Natural Reaction of Young and Aged Person with Telenoid in a Real World. JACIII 15, no. 5 (2011): 592-597.
  • [Ishiguro et al. 2016] Dylan F. Glas, Takashi Minato, Carlos T. Ishi, Tatsuya Kawahara, and Hiroshi Ishiguro. ERICA: The ERATO Intelligent Conversational Android. In Robot and Human Interactive Communication (RO-MAN), 2016 25th IEEE International Symposium on, pp. 22-29. IEEE, 2016.
  • [Yamanaka et al. 2006] Kazutoshi Takahashi, and Shinya Yamanaka. Induction of Pluripotent Stem Cells from Mouse Embryonic and Adult Fibroblast Cultures by Defined Factors. cell 126, no. 4 (2006): 663-676.
  • [Kawahara et al. 2017] Divesh Lala, Pierrick Milhorat, Koji Inoue, Masanari Ishida, Katsuya Takanashi and Tatsuya Kawahara. Attentive Listening System with Backchanneling, Response Generation and Flexible Turn-taking. In Proc. SIGdial, pp.127–136, 2017.
  • [Fujie et al. 2012] Shinya Fujie, Yoichi Matsuyama, Hikaru Taniyama, and Tetsunori Kobayashi, Conversation Robot Participating in and Promoting Human-Human Communication, The Transactions of The Institute of Electronics, Information and Communication Engineering (IEICE) A, Vol.J95-A No.1, pp37-45, 2012.
  • [Matsuyama et al. 2015] Yoichi Matsuyama, Iwao Akiba, Shinya Fujie and Tetsunori Kobayashi, Four-Participant Group Conversation: A Facilitation Robot Controlling Engagement Density As the Fourth Participant, Journal of Computer Speech and Language, 2015. (DOI:10.1016/j.csl.2014.12.001)


  1. Pingback: Video Friday: Teaching a Robot to Pick Up a Knife, and More - IEEE Spectrum

  2. Extremely interesting discussion. Do you think the philosophy and acceptance of robots in society would be accepted outside Japan?
    I have written a book which encorporates some of the sbjects discussed. The title is, Union of Opposites. It is on Amazon.

    • Yoichi Matsuyama

      Thank you for your comment! I just checked out your book on Amazon – it’s a very interesting subject. Professor Ishiguro and I also wondered the point during the discussion.

      Mary Upton, “Union of Opposites: An android’s quest to live and love in a human world” (
      Abstract: A12 is an android, created in Japan, in female form. She is given an artificial mind, allowing her to develop beyond any level of human intelligence currently in existence. Sold on to a large food company in Tokyo, she is befriended by her works director. Through him, she adopts the name Mirai Iwate. An attractive android, Mirai wishes to live alongside humans and be part of their society and future. She nurtures a budding ambition to travel and be involved in the American space industry. To do this, under cover, Mirai creates a fake identity in the name of Sarah Miles. WILL SHE BE WELCOMED – OR FEARED? Union of Opposites takes the reader on A12’s extraordinary journey from the laboratory, right into the pumping heart of humanity, before she fulfils her dream…

Leave a Comment

Your email address will not be published.