December Goal Post

The year is going by so quickly, I really can't wrap my head around the fact that it's already DECEMBER. But this month is very exciting as we have our January event soon approaching, so I guess the fact that it's already December isn't a completely terrible thing.

For the half-year symposium, my partner is Ved Kumar who is studying robotics. You can view his site and thoughts here. In meeting with Ved and discussing possible parallels one could draw between our two projects (I have expanded my part to be about neuroscience in general since C. elegans might be a little too specific), we eventually came to a point where we talked about the relationship between machine learning and human thought. As Ved's project trajectory has taken a new path towards an exploration of the ethics involved, he talked quite extensively about how machine learning - the ability of a robot or device to learn something on it's own - fits the scheme of ethical dilemmas in robotics because essentially, with the possibility of machine learning, robots would obtain a remarkably human ability. As such, the closer robot technology comes to the ways and characteristics of humans, the more people may (or may not) feel as though these technological advances have come too far.

Although we are still thinking about the product we will produce and display for our presentation (we were talking about possibly doing something with Ved's robot and my plastic brain, but nothing is really set in stone yet), our presentation will revolve around the following question: How does machine learning compare to human thought? As with any large question, there are other implicit inquiries within this one. By exploring a comparison of the two things, we can answer a more pressing question that probably has quite a few people concerned - Can machine learning become an equivalent to human thought? The thing with this however, as Ved and I continually returned to, is that human thought has an infinite amount of non-quantifiable values. This being said, I anticipate that the first thing required of our presentation is to clearly and explicitly define both human thought and machine learning so that the resulting narrative can also be clear and explicit. So the way we imagine this is that obviously, Ved would provide the expert information on robotics, machine learning, and some specific examples of the ethical arguments and I would talk about human thought. Based upon the nature of my work over the past two years, I am anticipating that I will take a relatively clinical route grounded in system biology.

At the same time though, I know that there needs to be an acknowledgement of those intangibles and the limitations they place on the possible comparison between machine learning and human thought. This then brings up a new question - What is the limit (if it exists) to robotic infringement on human nature's characteristics?

Another thing we touched upon in our discussion was the impact of disease on human thought and whether there could be an equivalent in robotics, beyond of course your occasional programming glitch. Although we did not choose to run with this idea, it was an interesting thing to consider as humans are always subject to the possibility of genetic error and thus disease. I wonder how, if at all, machine learning could change that? Does it imply that there must be a foolproof baseline code within the robot before it can begin to learn something on its own?

My favorite thing about this collaboration assignment is probably how having two people, two brains, and two projects involved, it becomes so much easier to see the questioning process at work and to observe the way in which one idea leads to another... times two. I also like how there is so much similarity, at the root of each of our projects and I think a few others will agree with me when I say that this complexity is certainly shining through in our collaboration assignment.

Besides the January day coming up, I'm working on completing the illustrations for my picture book and I'm also working on compiling soundbites for my second podcast. Over the winter break, I want to gather all my data and results, analyze them in the weeks following, and prepare the results section for my committee to review by the end of January/beginning of February. That's pretty much what I have planned for the next couple months. But as always, things usually don't go as planned so whatever happens, I guess we'll cross that bridge when we get to it.