Japanese robot likes sushi, fears president
Story last modified Tue Jun 05 08:09:52 PDT 2007
Kansei frowns when he hears the word “bomb,” smiles at “sushi” and looks scared and disgusted when someone says “president”–and he isn’t even human.
Japan’s latest robot, called Kansei and created by a university research team, can pull up to 36 different facial expressions based on a program which creates word associations from a self-updating online database of 500,000 keywords.
The English keywords then trigger the most appropriate facial expression, which ranges from happiness to sadness, anger and fear.
“What we are trying to do here is to create a flow of consciousness in robots so that they can make the relevant facial expressions,” said project leader Junichi Takeno, a professor at Meiji University’s School of Science and Technology.
“I believe that’s going to be a key to improving communication between humans and robots,” he said.
The robot has 19 movable parts underneath the silicone face mask. When the robot hears the word “president”, the online database picks up associated words such as “Bush,” “war” and “Iraq” and creates an expression which the researchers said is meant to mix fear and disgust.
Takeno says that in a few years, Kansei will also have speech abilities and will be able to convey feelings, which could be useful in places such as nursing homes for the elderly.
With Japan’s population expected to slide by around a quarter by 2050, and immigration a sensitive issue, some laboratories have developed humanoid robots that can work as maids.
Earlier this year, a university researcher created a robot that looks and moves exactly like him.
I think this is an interesting way to associate words and concepts outside of human consciousness. I wonder can this be built upon and construed into another application — sort of like an emotional search engine. Of the tip of my brain comes the ideas of people building mental models of themselves and then content could be grabbed based upon that model. It’s not necessarily based on known interests which would be reverse looking, but rather such a model could be predictive (and I’m thinking purely in terms of marketing to people individually) and dynamic throughout a person’s life. I think with this sort of thing you are looking for extreme triggers — things that make people very happy and things that upset people very much — because both emotional extremes propel action (whereas indifference spurs nothing).
What do you think?