robot emotions piece

In response to the question "Do you think robots will ever be able to feel human emotions?" my answer is a resounding yes. As a scientist studying the brain, the first question I ask myself along this topic is "what are emotions and how do they impact the brain?" While it's difficult to objectively quantify such things (emotional studies may use a scale to try and compare within individuals as one way to counter that issue), it's safe to say that brain areas are linked with emotional states. A study by Phan et al looked at this and found that certain areas respond to certain aspects of emotions. (http://www.ncbi.nlm.nih.gov/pubmed/12030820) Knowing that certain areas respond to different aspects of emotions (visual stimuli that trigger an emotional response influence behavior of the visual cortical areas for example) allows us to put that into simple "yes/no" points that could be inserted in a "if/else" portion of coding that could theoretically be altered by the robot itself upon learning.

There is a breadth of research investigating motivation/learning/fear using animal models that do not have nearly the amount of neocortex we do (the giant amount of tissue covering most of our brain with the squiggly lines). While one may hesitate to call these emotions, they are individual feelings/information that occur through brain signals like the emotions we think of and examples of how neurons/brain areas respond give hope for eventual studies in more complex models (whether that is in humans/animals/modelling). Adding to our knowledge will help programmers better mimic our brains and program in emotions.

The other avenue of research we have is through the disease state of people as a disorder disrupts their emotional states. People with bipolar disorder may be good candidates to see the different brain states compared to normal age-matched samples. Other disorders affect different areas of the brain and research can be done to see what, if anything, is affected and what is preserved in order to better isolate how emotions are affected by disruptions from normal brains.

While it sounds complicated (believe me, you truly have no idea how amazing and yet frustrating our brains can be), I have faith that we'll one day be able to replicate emotions through these (and other) types of studies which will only help our knowledge base and programming partners to better improve AI. I won't say it's anytime soon or even within our lifetime as the progression of science is hard to put on a simple timetable, but I do feel it'll occur at some point.