Voice and facial recognition could help AI surpass humans in emotional intelligence [Mikko Alasaarela, Inbot | VentureBeat]
While humans are still struggling to understand each other, emotionally intelligent AI has advanced rapidly.
Cameras in phones are ubiquitous, and face-tracking software is already advanced enough to analyze the smallest details of our facial expressions. The most advanced cameras are able to tell fake emotions from the real thing in certain cases. Voice recognition and natural language processing algorithms are also getting better at figuring out our sentiment and emotional state from audio.
An example of assuming we can identify emotions through facial recognition… but the theory of constructed emotion says this is wrong. There aren’t hard-wired basic emotions that are inevitably associated with one or another facial expression. The same expression could indicate anger, excitement, or surprise, for example.
This is a far too simplistic understanding of emotional intelligence:
The problem is that only 35% of men are emotionally intelligent.
You ever wonder why I tell women to be the CEO and treat men like interns? That’s why. 65% of guys are incapable of being good partners.
Intelligence is usually measured on what scale developers intend to be interpreted as an interval scale, as a continuum of numeric variables that are ordered and in which the differences between each scale point is considered to be equal. This would certainly be the case for most operationalizations of emotional intelligence. So saying someone has or doesn’t have emotional intelligence in binary fashion makes little sense.
Katz got that 35% figure from who he calls the “Godfather of Couples Counseling,” John Gottman. Gottman is famous for claiming he can predict which couples will and will not divorce simply by watching them discuss a contentious topic for fifteen minutes. Unfortunately, his use of predictive modeling was entirely flawed.
What Gottman actually found was that 65% of men increased negativity during a conflict (as observed in Gottman’s research) which is not the same as saying “only 35% of men are emotionally intelligent.”
Woebot: AI for mental health [Andrew Ng | Medium]
I’m thrilled to announce I am joining Woebot’s board of directors as its Chairman. I will be assisting its CEO, Alison Darcy, and the company in its mission to build a chatbot that will help the millions of people who struggle with their mental health.
Woebot is based on cognitive behavior therapy. In an RCT, Woebot reduced symptoms of depression and anxiety in two weeks. Efforts like this go back to the mid-1960s, when Joseph Weizenbaum developed ELIZA at MIT.