Commercial robotics research in Japan is essentially an inevitability as the population there ages and as technology improves. In her 2006 paper, Kayoko Ishii suggests that cognitive scientists should piggyback on this inevitability and use these robots as an opportunity to deepen our understanding of human thought.
She cites a definition of computational neuroscience as “to investigate information processing of the brain to the extent that artificial machines, either computer programs or robots, can solve the same computational problems as the brain, essentially in the same principle.” If you buy into this definition, then computational neuroscience research in cognitive robots makes a lot of sense.
Specifically, she notes that the human brain consumes 20% of body’s energy, and a lot of this is expended in social interactions. A robot would have to have a large amount of information processing in order to keep up with humans on a social level. Also, robots must have the ability to autonomously change their own algorithms based on their interactions with the environment, just like humans do.
Currently, the infant-like robot “Infanoid” can conduct “eye contact” by taking video images from its camera and recognizing when and where the human is looking. It can also conduct “joint attention,” by looking at a place where its cameras recognize that a human is pointing.
The more of these types of activities that a robot is able to conduct, the more us humans will consider them human-like. Since humans interpret the world as acting intentionally, humans will project emotions onto robots even if they don’t have any.
Many people make the vague argument that “robots will never achieve consciousness,” an argument that Ishii deconstruct this argument thoroughly, claiming that it stems in large parts on Western religious ideals that Japan does not share. I would add that just because we have not yet formulated detailed examinations of self-awareness or other “uniquely human” functions does not signify that we never will. I will hear nothing of limitations based on Godel’s complexity theorem, as that argument has also been sufficiently debunked.
Ishii, K 2006. Cognitive Robotics to Understand Human Beings. Quarterly Review: Science and Technology Trends 20 (Pdf here).