The Research Loop
How’s your research going?
A pretty valid question, given that I’ve been holed up since November working on it. But it is kind of hard to answer if you don’t know how I do research.
I’m designing a new algorithm. I’m not going to talk about what it does at the moment, but how I go about doing my work. Some people may draft a plan, and then tick off the activities on the plan one after the other until they are done. What I do is much more messy than that. I think of it as experimental algorithm design. To me, research is a loop:
1. Try something. 2. It doesn’t work. 3. Find out why. 4. Figure out how to get around the problem. 5. Go to step 1.
I’ve gone around this loop more times than I can remember. The important parts are steps 3 and 4. Nothing I’ve done since November has worked, but that doesn’t mean I’m not making progress. A reporter famously asked Edison about his failures in trying to make the light bulb, and he said he’d learned a thousand ways to not make a light bulb. I really feel like that most days.
It is supposed to be hard. I’m working on a problem that many others have worked on. I’ve been to the library, and copied stacks of journal articles of what everyone else did. In my gut I don’t think they’ve found the best solution that is out there. To date, there isn’t even a very good solution.
The trick is to not let step 2 get you down. All my brilliant ideas from step 4 just get slammed into the dust as I loop around to the beginning. I find myself procrastinating after I think I know how to fix a problem, just so my mind can have a little happiness before finding out it was all wrong. The emotional drain of failure is the hardest part.
Some day, step 2 will work, and I’ll be done. I really hope that day comes soon, but it is hard to tell.
