- Time: 13:30- July 26th, 2015
- Place: Lecture Room 1, Information Science Building, Nagoya University.
(A4(3) in the map: Nagoya University Higashiyama Campus Map) - Program
- 13:30-14:30, Kojiro Honda, TBA
- 14:30-15:30, Tadahiro Taniguchi, ``Symbol emergence in robotics: constructive approach towards human cognitive and semiotic systems''
- Break
- 15:45-16:30, Minao Kukita, ``Morality, emotions, and ethical symbol grounding''
- 16:30-18:00, Wendell Wallach, ``From Moral Machines to A Dangerous Master''
No registration is required.
Abstracts
- Wendell Wallach, ``From Moral Machines to A Dangerous Masters''
How can the emerging societal challenges posed by AI and robotics be managed through ethics, engineering, and oversight? From the initial mapping of the new field of enquiry known as machine ethics or machine morality, to recent breakthroughs in machine perception and machine learning, the controllability and safety of artificial intelligence has taken on increasing importance. Deep learning networks capable of assimilating vast databases are laying a new foundation for future work in artificial intelligence and robotics that is likely to be quite beneficial but also carries risks. Alert to those risks, leading scientists have suggested new research trajectories to ensure that AI will be robust, safe, and demonstrably beneficial. This talk will trace those changes and new research and policy trajectories. I will draw upon themes in the 2009 book Moral Machines:Teaching Robots Right From Wrong, which I co-authored (with Colin Allen) and from my recently published A Dangerous Master: How to Keep Technology from Slipping Beyond our Control (Basic Books, June 2015).
- Tadahiro Taniguchi, ``Symbol emergence in robotics: constructive approach towards human cognitive and semiotic systems''
Humans can acquire language through physical interaction with their environment and semiotic interaction with other people. To understand computationally how humans can form a symbol system and obtain semiotic skills through their autonomous mental development is very important problem. In the research field of symbol emergence in robotics, we are challenging to construct a robotic system and machine learning method which can obtain language through embodied multimodal interaction. In this talk, I introduce the basis of our research and state-of-art results. Specifically, I talk about multimodal categorization and double articulation analyzer which enable a robot to obtain words and their embodied meanings, in unsupervised way.
- Kojiro Honda, ``Developing a ‘Charter of Roboethics’: From a Retrospective to a Proactive Approach in Robot Technology Engineering''
In this paper, the current condition of robotic development in Japan will be introduced with indicating academic roadmap for 2050 which was prepared by Japanese roboticists. If their future vision would be realized, then we have to exist together with robots in diverse ways. The Situation of Japanese Engineering Ethics: It is about 10 years ago that engineering ethics was imported into Japan. Now it becomes one of standard courses in technical faculty in Japanese university. Thinking ethical issues of technology habituates in technological world in Japan. However, in the course, the main topic is analysis of failure cases and consideration for the prevention of the same failure. This can be called ‘retrospective approach.’ There is little opportunity for technical students to think about future technology and its social consequence.
The Future Vision of Japanese Roboticists: In such a situation, The Robotics Society of Japan, The Japanese Society for AI, and Japan Ergonomics Society collaborated to make an academic roadmap of robotic development (2007). Japanese engineer’s realistic view of future includes following (A-G).
- A) Self-reference, self-evaluation, self-restoration system: Development of the robots’ ability of “Meta-Cognition” to understand own state and own cognitive contents, then, to modify own action.
- B) Cognitive development system: Development of the system which can spontaneously extend its own cognitive faculty via interaction with outwards and others.
- C) Comprehension of human objective and intention: Development of robots' ability to understand others' intention and objective for making a good communication with people.
- D) Autonomy: Development of the robots’ ability of self-decided action without outer command.
- E) BCI (Brain-Computer Interface)/BMI (Brain-Machine Interface)/Cyborg: Realization of combination between machine and human being, which could not only support our ability, but also extend it.
- F) Artificial body organs: Development of human-made body parts.
- G) Wet robots: Development of robots made of biological materials.
They think that these technologies should be developed by 2050. We can imagine many socio-ethical issues (as following I-V) will happen according to the development in near future.
- I. Ascription of Responsibility: Who will have the responsibility of robot's accident?
- II. New type of Private Information: How can we protect a lot of users' private information which robots gather?
- III. Educational Effect for People: Children might imitate humanoids as adults. What will be the consequence of it?
- IV. Estate of Robots in our society: Should robots be treated as quasi-human?
- V. Human Enhancement: What is the consequence of body-robot fusion in the future?
It is desirable for public to discuss these issues before engineers finish the development. Discussing issues in advance doesn't mean regulation of technology. The result of the discussion would be fed back to engineers. And that could be helpful for them to know the specific feature of the socially acceptable robots.
The Start-up of Roboethics Studies in Japan: As mentioned above, Japanese engineering ethics education is based on retrospective approach. This approach is not suitable to the assessment of future technology. We need a way to anticipate the social effect of technology that could be called as `proactive approach.'
The decision of moral acceptancy of a new technology is dependent not only to the normative consciousness of engineers, but also to public sense of moral issues attached to the technology. Proactive approach can give the both social actors a platform where they communicate and share the moral issues. And philosophers could have a special role to make such a platform.
To be concrete, philosophers can provide ethics charter for future technology to give a material for discussion between engineers and public. As for robotics, it is useful to make a `Charter of Roboethics.' In 2011, some Japanese philosophers in younger generation gathered, and started to discuss the moral issues about robotics. Now they are making the charter of roboethics in Japan.
This paper will introduce the skeleton of the coming charter.
Prototype for Charter of Roboethics: The developing charter has 3 fundamental ideals as following.
- (1) Human Security: Physical Security and Psychological Security
- (2) Enhancing the community by Robotics: Human Enhancement -> Human Re-enablement
- (3) Anti-Military: Robotics for Peace
And it will include at least 4 articles of general issues as following.
- 1. We use robots for morally clean purpose.
- 2. We use robots to promote working and life environment.
- 3. We don't develop robots whose purpose is doing harm to people.
- 4. We respect the tradition of our own society. And our purpose of developing robots is not changing but enhancing our society.
At the present stage, we have 3 must-be-included issues as following.
- A) If some robot's act would have morally important consequences, we should make clear the locus of responsibility before using the robot.
- B) We must respect the users' private information (PI) that robots have gathered. If we avail of the PI, fully informed consent should be given.
- C) Roboticists should give consideration to the robots' psychological effectfor users.
The process of making the charter is ongoing. We will have more articles in this charter. The completed charter will be open to the society as a material for discussion about moral issues concerned with robotics.
- Minao Kukita, ``Morality, emotions, and ethical symbol grounding''
Some researchers in artificial intelligence and robotics have recently attempted to simulate our moral decision making and moral actions by artificial systems. Their motivation is primarily practical, for example, the concern about potential disasters caused by complex interactions among machines, software agents and humans. However, to create artificial moral agents is also of great theoretical interest. Just as artificila intelligence of the last century inspired philosophical debates over what intelligence is and deepened our understanding thereof, artificial morality may have a significant impact on our debate over and understanding of what morality is. In this talk, I will focus on a theoretical question concerning morality in the machine. Specifically, I will address the ethical counterpart of the symbolgrounding problem, i.e., how artificial system can understand human values. This question also has practical importance, because
About the speakers
- Wendell Wallach is a consultant, ethicist, and scholar at Yale University's Interdisciplinary Center for Bioethics. He studys ethical issues concerning emerging technologies. He co-authored (with Colin Allen) Moral Machines: Teaching Robots Right from Wrong (2009). Recently he published Dangerous Master: How to Keep Technology from Slipping beyond Our Control (2015).
- Tadahiro Taniguchi is an associate professor at Department of Human and Computer Intelligence. He advocate the concept of ``symbol emergence system'' that supports human linguistic and semiotic communication through constructing robots and artificial intelligence. He is the author of 『コミュニケーションするロボットは創れるか―記号創発システムへの構成論的アプローチ Is Communication Robot Possible?: Constructive Approach to Symbol Emergence Systems 』 (2010) and 『イラストで学ぶ 人工知能概論 Introduction to Artificial Intelligence (Learning through Illustration)』 (2014).
- Kojiro Honda is a lecturer at Department of General Education, Kanazawa Medical Unviersity.
- Minao Kukita is an associate professor at Graduate School of Information Science, Nagoya Unviersity.
Contact Information
Minao Kukita (久木田水生), minao.kukita@is.nagoya-u.ac.jp