There are a lot of articles today about how a Japanese company has created a robot that senses people's emotions. But this isn't the first research in this vein. I've seen a lot of talk about machines recognizing the emotions of users. Academically, is an interesting challenge. But I've never understood the practical purpose of it. That may surprise you, given technology's cold and impersonal nature.
Yes, I'm sure that one day it will be very useful for machines to notice or emotions. For instance, consider a machine you can give verbal commands to in plain English. Since human language often has subtleties undertones, it would be quite necessary for it to take notice of your emotional state.
But for today's machines, we're lucky if Siri can make out or words, never mind pick up an emotional subtext. I’ve never been able to think of a time when I would want an electronic device to behave differently depending on my mood. Sure, I’ve wanted it to become easier and more reliable when I’m angry, but that’s not too likely.
For the most part, we’re still at a point where our devices obey specific instructions. As long as that’s true, we want them to be predictable. Introducing some changes based on its perception of our emotions is going to be annoying, even if its perception is accurate. Once we get to software that is somewhat autonomous, it might have to make decisions and judgements based on circumstances, including the mood of the user. But such software is still a challenge in itself, so there’s no immediate need to sensing emotions.
No comments:
Post a Comment