The Powers of Supercomputers

I was really excited when Prof. Petit brought up IBM’s Watson in class as I had read about before. The discussion we had in class was definitely insightful.

At its core, Watson is one of the most advanced simulations of artificial intelligence that exists. When you think about the amount of programming and algorithms involved to make Watson comprehend and answer Jeopardy questions, it’s really a big surprise. I know I take the English language for granted (I guess any language will pose the same problems for Watson though), but learning about this made me appreciate just how complex our minds are.

We discussed potentials uses of supercomputers like this in the future. I’m not sure if Prof. Petit said this based on a hunch or if he read actual data, but he mentioned using supercomputers like Watson to help diagnose illnesses. The other day, I found an article on this very idea from Columbia University (see link below). Herbert Chase of the Department of Biomedical Informatics (wow that sounds interesting) mentioned that Watson’s powers would come in handy for developing custom-designed treatment plans that take into account all of patient’s data, and can also help doctors answer unsolved questions.

http://news.columbia.edu/research/2360

One obvious note he made was that Watson will never be able to read the signals (or emotions) coming from a patient. However, I do think this can be possible ONE day. Everything can be broken down into a science, right? Language is a science of semantics and syntax and so on. Language is an accumulation of rules along with contextual cues that help eliminate ambiguity. Much like they did with Watson, couldn’t they break down all 3,000 (just throwing a random number out there) facial expressions into a series of measured movements of the eyebrows, lips, eyes, forehead, cheeks, and nose? From there, they’d have to encode what combinations mean what. Then, they’d have to analyze enough different faces to be able to encounter variances. It might require a lot more processing power than Watson, but by whenever this is ready, I’m sure the technology will exist. Never able to read patient’s emotions....YEA RIGHT. I’m the kind of person that believes technology has no limits.

A concern however, is then what happens when u have the technology to fully read human emotions? Well you slap that technology into a Honda robot then ship them out into society so we can finally be friends with robots!