0

It took a while before the research and discussion around human intelligence evolved from the basic IQ based testing and evaluation to what is now a multi-dimensional view of intelligence from aspects such as EQ (emotional quotient) and SQ (spiritual quotient).

While there are probably even more dimensions of intelligence that I can't even name but worth thinking about, I have wondered how these different dimensions have been applied or implemented in the design of conversational interfaces or chatbots.

We would expect that an AI would be able to solve complex IQ related issues much more proficiently than humans could, but in terms of being able to engage with people on an emotional or spiritual level, what does it mean to be able to design the user experience around other dimensions of intelligence? Do people really expect artificial intelligence to have these other dimension or characteristics of intelligence?

My question is: what is the research around people's expectations of articial intelligence? Is there a similar barrier against an artificial intelligence feeling too realistic, the same way that the aesthetics of robots that are too life-like can cause some negative or uneasy experiences (i.e. the 'uncanny valley')?

UPDATE: the concept of a digital avatar isn't all that new apparently, and these are just some of the references that I have seen on the concept recently: http://www.idavatars.com/ http://www.foxnews.com/tech/2017/02/01/why-youll-send-digital-avatar-to-meeting-by-2020.html

2

It seems like much of the research on expectations that humans have for robots is actually rooted in appearance as opposed to specific behavioural or emotional characteristics (though you could make the case that things like naturalistic movements and facial reactions are approximating emotional characteristics).

But more specifically to the point of the Uncanny Valley - In doing some research on this question myself, I found that it's actually a much more contentious theory than accepted fact. This research (opens a PDF) states that the uncanny is a result of different design aesthetics instead of any inherent issue or line that exists when one gets too close but not quite to human status.

But, to answer your question, the research doesn't seem to exist yet, or at least not along the dimensions that you're curious about. Most of it relates to physical characteristics and visual responses to stimuli.

2
  • +1 A good summary of the 'uncanny valley' and reference, just wondering if you have any personal thoughts about it yourself?
    – Michael Lai
    May 8 '17 at 12:11
  • 1
    Yes! I personally think the uncanny valley re: behavior or emotion is possibly more binary than it is for appearance. It's the Turing test, right? Either the computer passes as human or not. I sincerely doubt there's an almost-human level of AI that causes the revulsion/dislike that robots show visually. I'd love, however, to study the chat bots that spammers use on dating sites to drive traffic to their sites - I bet that data set is fascinating.
    – kristinalustig
    May 8 '17 at 17:16

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.