We'll likely see a major turning point when chatbots can communicate via animated human face images instead of just text. Right now chatbots seem largely limited within what we might call the "nerd community". When the interface changes from text to human face a much broader segment of the public is likely to become involved, and most of these folks will not be reading intelligent informed articles about AI, such as you have here. So, many more people, who are far less informed. Sounds like a problem multiplier to me.
Old software can already animate a face image from text or audio input. But all you can do is tell the face what to say, it's not interactive. Making it interactive would seem to mostly be a matter of processing power. I have no idea how much is needed, or how close we might be to a convincing human face interface.
Interesting idea. I wonder if we hit the "uncanny valley" problem here though where the more lifelike an animation is, the more uneasy it makes people. An alternative may be more cartoonish interfaces -- although most people are very savvy with text interfaces (think texting, twitter, Google searches etc)
My thought was that a human face interface will be a far more compelling experience to far more people, and that will multiply both the pluses and minuses. For neglected people in nursing homes, it might be great. For those with a weak hold on reality, maybe not so much.
I just finished an article (unpublished) that asks this. We select our human contacts based on how well they can provide whatever experience we are seeking. Why should I talk to you or you to me if an AI bot could meet our needs better?
When chatbots have a text interface the boundary between real and software is still fairly clear. As the interface becomes more realistic looking, that boundary may start dissolving in many people's minds.
We'll likely see a major turning point when chatbots can communicate via animated human face images instead of just text. Right now chatbots seem largely limited within what we might call the "nerd community". When the interface changes from text to human face a much broader segment of the public is likely to become involved, and most of these folks will not be reading intelligent informed articles about AI, such as you have here. So, many more people, who are far less informed. Sounds like a problem multiplier to me.
Old software can already animate a face image from text or audio input. But all you can do is tell the face what to say, it's not interactive. Making it interactive would seem to mostly be a matter of processing power. I have no idea how much is needed, or how close we might be to a convincing human face interface.
Interesting idea. I wonder if we hit the "uncanny valley" problem here though where the more lifelike an animation is, the more uneasy it makes people. An alternative may be more cartoonish interfaces -- although most people are very savvy with text interfaces (think texting, twitter, Google searches etc)
My thought was that a human face interface will be a far more compelling experience to far more people, and that will multiply both the pluses and minuses. For neglected people in nursing homes, it might be great. For those with a weak hold on reality, maybe not so much.
I just finished an article (unpublished) that asks this. We select our human contacts based on how well they can provide whatever experience we are seeking. Why should I talk to you or you to me if an AI bot could meet our needs better?
When chatbots have a text interface the boundary between real and software is still fairly clear. As the interface becomes more realistic looking, that boundary may start dissolving in many people's minds.