Sunday, October 18, 2009

The unasked question

I'd like to start by apologizing for the lack of a post last week. It probably wasn't too disappointing for most of you, given that we had a chance to catch up in person, but for those of you whom I didn't get a chance to see for one reason or another (JB, AO, RL, and GS, to name a few), I'm sorry. I really don't have the time now to do justice to the trip in a full write-up, so I will leave it at this: the time I spent back in Chicago was fantastic, and my only regret is that I couldn't stay for longer. My thanks to you all for your hospitality and warmth- YJ felt very welcome as a result.

After I returned Sunday morning, I spent the day gleefully assembling my new robot, a picture of which is embedded below.
As is evident from the picture, it is not a very expensive arm, but it should let me practice my new-found robot skills. The rest of the weekend was spent taking care of the usual perfunctory tasks and visiting with my immediate family. The remainder of the week was unassuming, if short, and the weekend was spent doing work or getting bored of that and playing Scribblenauts. Yes, at KB's suggestion, I picked up a copy of Scribblenauts, a game in which you can create just about anything in order to solve puzzles. Ornithopter? Yep. Keyboard cat? You bet. Cthulu? Naturally. Don't let the cutesy, cartoonish exterior fool you- this is a worthwhile game for the DS. At one point, I was curious to see if it would recognize the word "atheist." It accepted it, and out popped some guy with white hair. Intrigued, I typed in "god." A Zeus-like being came into being, which frightened the atheist, who ran and hid on top of a lampshade.

Registration for classes begins tomorrow here at Purdue, and before I can register I am required to submit a plan of study, which is an outline of what classes I plan to take and when I intend to take them. This was always a non-occasion for me at IIT, with the exception of the assault on the hapless servers come midnight*, because I had always had my classes charted out well in advance. My choices were obvious, as they were either the required classes for my major or professional electives which I filled with anything robotics related.

Now, however, I have met putting together a plan of study with some hesitation rather than swift action. This puzzled me; surely I know what I want to do, right? For some long time now, I have answered the question of what I want to do in life with a simple refrain: "Build robots!" However, I am now posed with the unasked question, "Specifically, what?" The old answer no longer works, as it is too broad to determine how I should spend the next 2+ years of my life. I had seriously considered working on multi-robot systems, which are often analogized to ants working together to complete a task, but I have little experience in this area and as a result cannot put my weight fully behind it. Moreover, my true intent is to build robots that help the general populace, that take care of the mundane flotsam which occludes daily life. Do swarm robots work towards this goal? Possibly, but there is much more that needs to be present. An orderly squad of ants is useless if they don't have the skills required to accomplish a task.

This discussion monologue hasn't answered the question of how I want to get my degree either. Should it be done primarily through taking classes, with research as a small aside, or should it be done with a focus on research? The typical answer is research, but I'm not certain that it is the proper route for me. Most professors will not take Master's students to be paid Research Assistants, as they want PhD students whom will be around for much longer, and I don't think that I can commit myself to seeking a doctorate at this time.

In conclusion (finally- I've been a bit long-winded :/ ), my path seems a bit more clear after this bit of semi-public introspection. I should focus on just taking classes for now, keeping an eye open to possible research opportunities but not banking on them. As for what classes to take, I plan to focus on classes regarding AI, machine learning, and machine perception. In non-Jeff speak, that means learning how to make robots understand what is going on around them and decide what to do from there. Others have done quite a lot of work in this area, and I think it would be more fitting for me to learn from them first before charging straight into research armed with nothing but my undergraduate skill set. Because it is somewhat unconventional, however, I am left wondering if I am making the proper decision. Ostensibly it seems so, but I am not yet certain.

-JM

*: Ah, the fond memories of the registration races are coming flooding back. The shouts of "F5!" heard down the hallway, the agony of having your session ended, the smug looks on the faces of those who managed to get through.

No comments:

Post a Comment

Thanks for commenting! As this is among friends, comments are not moderated. However, please refrain from using people's full names (besides your own) and the name of my employer. Instead, give names as initials and my employer as [redacted].

I read all comments immediately, but may not reply right away. I will reply with a comment, and send you an email with my response; the email may contain some private content as indicated by italics.