ProgramsView all programs
From the news
- Chancellor Amma Addresses the Parliament of World’s Religions
- Amrita Students Qualify for the European Mars Rover Challenge
For the researchers of AMMACHI Labs and CWEGE to be exposed to the high calibre of research and work that has done, and to gain insights into the research process through [his/her] experiences. Additionally, any potential collaboration that comes out of these interactions is welcomed.
Participants: AL & CWEGE Research Staff & Scholars
Dr. Peter Gardenfors has been a Senior Professor of Cognitive Science at Lund University since 2016. He has also been a Senior Research Associate with the Palaeo-Research Institute at the University of Johannesburg since 2019.Among his numerous awards and prizes are: The Rausing Prize in humanities in 1996. The Prize for interdisciplinary research by Academiæ Regiæ Scientiarum Upsaliensis in 2008. The Hermann Lotze Prize in 2012. Universitatis Lodziensis Amico Medal 2012. Fellow of the Cognitive Science Society since 2016. Awarded the Einar Hansen Honorary Prize in Humanities 2017.
Published more than 400 articles (see http://lucs.lu.se/people/peter-gardenfors-bibliography-2010-present/ and http://lucs.lu.se/people/peter-gardenfors-bibliography-2000/). The journals include Science and are within philosophy, cognitive science, logic, artificial intelligence, economics, management science, linguistics, psychology, archaeology, anthropology, sociology and biology.He has more than 23,000 Google citations, h-index 57, i10-index 159. The most cited scholar in the humanities in Sweden
The event start with what we think about robots is we are kind of spoiled with pictures about robots from movies and just collected some few pictures and this all look like human and I don’t really look humans and if you look around you don’t see these robots I mean they don’t they don’t exist and they will not exist in a very long time probably never because that kind of construction we see here is not a very good one so what kind of robots do we see okay we see washing machine as an household robot is washing machine that’s a machine and it runs its own and it cuts your grass that’s all fine and when you have a vacuums cleaner all automatic they look like they are not very social they are about the socialist cockroaches and they behave like cockroaches I mean this vaccum cleaner hides under the beds and so on so we don’t really interact with them but you don’t see anybody talking to a lawn mower or having fun with it maybe we can trick it but that’s another story so these are the kinds of robots we are seeing at the moment that exist in our environment there is another kind of robots coming up there are experiments going on with autonomous cars and they are not really on the streets yet they are all kinds of legal problems putting them upon on the streets but the technology is they can ruin and run quite safe even if it’s never perfectly safe I would say that riding is an autonomous car would probably be safer than running in a car driven by a human there are some experiment things I have here picture an autonomous bus that running a regular line in Stockholm I am not sure that that this is the first regular line like by an automats vehicle but it still runs on an experimental basis. There are robots that interact with the people I mean a car into some extent will have to interact you have to tell the car where you want to go But there are experimental social robots there is a Japanese robot desired to help people with different kinds of handicaps so it can serve a breakfast and it can solve some problems like a pouring a cup of coffee bringing out juice from the refrigerator but it does so extremely slowly and extremely clumsy so it’s not like a human helper but it’s still it’s an autonomous thing here is a German hospital robot its main task is to serve drink and medicines to the people in the hospital mainly elderly people and you can see on the face of the women here she is bored and the robot is not very interesting thing there are other robots like animals and icon robots.
Mind reading its nothing magical we define mind reading as an ability to understand what others feel want and think in philosophy this area is called theory of mind psychologists call it as intersubjectivity. So today am gone talk about how can we create a robot that has the capacities to understand what humans feel want and think and use this knowledge to improve its interaction with humans so the lawnmower and the vaccum cleaner they had definitely don’t have any capacities to read the minds of human cars down to either they have to ask human where he or she wants to go but that’s about the only mind reading that comes. There are three kinds of mind readers first are the human mind readers then the animal mind readers and then the robotic mind readers
Components of mind reading
● Understanding the emotion of others
● Understanding the attention of other
● Understanding the Intentions of others
● Understanding the Beliefs of others
● Understanding the mind of oneself
Understanding the emotions of empathy
● Empathy: perception of emotion in another activates the same emotion on the receiver
Evidence for empathy in mammals and birds
And there are some examples of kismet expressing emotions they use their ears I mean we can’t use our ears to express emotions but we know from animals and dogs the ears can be part of emotional expression and they have used this for this kismet robot and we cannot
identify the emotion on this robot so it’s a problem. Then he gave an example regarding the understanding of human attention a paradigmatic example the human says to robot. Bring me the cup”
If there are many cups in the environment the robot must disambiguate by using a mode of attention of the human.
Two main methods are
1. Following gaze or pointing
2. Modelling the current interest of human
Understanding attention of others
Children at 6 months can follow the gaze of their mother if she turned her head, at 12 months they can follow gaze of their mother if she just moves her eyes.
How does a robot control the attention of human?
The role of attention in language production
suppose you want to express that icub gives the ball to peter, peter is given the ball by the I cub, the ball is given to peter by I cub, it is given by the I cub to him or you can say peter receives the ball from the robot and we can choose various forms and receives here the action that distinguish between the general rule is the subject of the sentenc3 expresses the focus attention of the speaker. The choice of verb can focus on action (cause or a result). Manner verbs (push, walk, hit) express actions while result verb (move, fall, reach) express results. And in order to understand the intention of others language helps but is sometimes not sufficient. So the task for a mind reading robot is 1) develop human robot joint attention techniques 2) model a flexible system for reading intentions 3) model joint intentions 4) model joint beliefs ( common ground) 5) model morality.
Gayatri manikutty: do you see differences in the kind of joint attention area or in empathy then you work with children rather than adults because we see that there are various interactions that need to be designed and when you are working with children what were some of the more critical factors that really made an impression on the child which probably for an adult could be more on the functionality or the purpose rather than you know the interaction itself so I would like to get some of your thoughts on that aspect and also the other thing was that all the robots you showed had the embodiments of the robot the face is actually an actuated face with the eyebrows and with eyes actually moving and things like that so I wanted your thought on whether to go with the actuated face or not?
Speaker: I think go with an actuated face that’s very interesting almost all the robots you seen except for the lawn movers and vacuum cleaners had some form of facial structure and of course for humans that’s very important we have the eyes mouth these are main tools for the interaction so that’s our important natural tool and we are very skilful at using them difference between adults and humans in interaction and you do because in humans in adult we know we are interacting with somebody else that even if it seems that we have a common goal at me maybe that the other individual wants to exploit us want to use us want to fool us so we are aware we have this awareness that this person reliable or should I cooperate or should I not trust the person so we have the kind of second level of mind reading that I am aware that the other person may want to use my options or so on to get the benefit without me gaining anything and children they don’t have the capacity they don’t have understand false belief thoughts they think everybody needs to cooperate.
Devash: when you told about the idea of purpose or the intention feeding into a robot I was thinking you know when we call something of purpose or intention we are basically referring to a higher goal you know for example sweep the floor it would be a small goal when you are thinking about building an entire ice men it’s a larger goal that requires a bigger intention that will require a larger structure a larger organization and work distribution working with multiple people and which comes does to building consciousness into working together which is basically all human qualities but no other species have this quality so my question is there is a lot of research going on how to build a robot which comes close to being a human should the focus rather be on building machines or robots which are best at human interaction you know user friendly machines rather than building robots we should we should be targeting at building user friendly machines because to be honest robots will never be able to replace humans that’s what I feel I might be wrong please correct I think robots can replace humans because they have already done so in routine works I means in lots of industrial work putting together a car is done by welding robots and so on movements and they do them very quick and fast and I don’t have to take breaks I don’t have sleep and what not there have replaced already the boring and may be labourers in industrial works what I was thinking was when I talked about common intentions and so on was more ordinary situation working together in a workshop in a carpentry or in a washing room whatever where you do your things you robot is doing its things but you have to interact and you have to see what the other one is doing of course you may have building something and cooking or constructing a wooden chair or but there are many sub goals and when reading intentions it’s not just having this general goal that is important but you have to break it down to understand that now this person is going to fetch piece of wood to make the log of chair.
Chris: when I was doing some philosophy of mind studies in undergrad and a little bit grad school they always use the touring test as the prime example of how to tell the computer has you know the rudimentary form of consciousness and it was always you know as like the classic experiment and we have come so far since then but then when you were it struck me when you were talking about how we need drop this idea of making a human and how that has so many assumptions that come along with that are actually counterproductive so I am wondering if that makes touring test more irrelevant or if there some better way to test computers or robots so called consciousness?
Speaker: Well the classical touring test is that you have converse with somebody via a computer and if you can’t tell whether the one you are conversing with is a computer or human then the computer process or the computer program passes the Turing test so far chess this has been passed for long time ago playing chess you can’t tell whether well you can tell because the computer is much better than a human being but that’s another but the Turing test presumes that we use language that’s problem but I think you are perfectly right that having that as a test is not a good one I don’t really know how to answer the question you should use instead I mean I mentioned that most robots are boring you don’t want to interact with them for a long so I would say that robot that’s is interesting to interact with would be or maybe better way of formulating a touring test somebody that has not the same response all the time.
Attendance /No.of participants: 34
List of the participants