Confessions of a researchaholic

August 3, 2013


Filed under: Real — liyiwei @ 11:36 am

Through my open mentoring program I got to experience many students with diverse backgrounds and characteristics. Some of them can be quite strange. Here is a recent case study.

I asked this guy to implement a standard software system in computer graphics. Instead of doing that, he kept on reciting on a daily basis what he read in graphics textbooks, none of which is remotely related to the target task. I tried to clarify the assignment repeatedly but he could not seem to comprehend. Finally, he told me that he wanted to appear being “productive” to get my attention.

[He did get my attention, but only the negative kind. The result is a prompt termination of our external mentorship.]

By definition, productivity means things other people want and care about, such as research papers or industry products. Reciting what you have learned is not productivity – no one else cares about that, and it cannot be measured. It is actually more like anti-productivity – a waste of time for everyone either to write or to read.

On some level this is probably related to the fundamental distinction between study and work. Learning is necessary for eventual productivity, but it is not producing anything by itself. Instead, it is anti-productivity, sucking away time and resources that can be used for productive work.

This is a key reason why I only teach undergrad courses, because courses (especially non-seminar, non-project-oriented ones) may trap the grad students in the undergrad mentality. The right way to learn is by an active on-demand basis – seeking what is found to be necessary during the research work, instead of a passive batch process – reading entire books or taking entire courses without really knowing the relevance.


  1. I’m not convinced this is always the only “right way” to learn. You don’t always know what you don’t know, and if you only ever pursue learning different topics in a strictly on-demand fashion, you may remain blind to a whole branch of knowledge that could be of tremendous relevance to a given work project. Many interesting and creative scientific advances have arisen because ideas from diverse areas came together in surprising ways, which such a focused approach would discourage.

    Put another way, why is pure study appropriate/acceptable for undergraduates, but incorrect for graduate level research?
    It seems to me some mix is more appropriate overall, a balance of “explore vs. exploit”.

    Comment by Christopher Batty — August 3, 2013 @ 6:58 pm | Reply

    • Good point. I probably should have clarified the context a bit though. Basically, everything I did with Asian institutions (MSRA or HKU) has to be biased against the prevailing culture and practice.

      For example, within the context of this particular post, the Asian education system is already heavy on the passive side. So I need to be more on the other side of the extreme.

      If I were in Stanford now, I would opt teaching grad courses.

      I also expect grad students (not just in Asia), after being “fed” as undergrads, ought to have a broad enough knowledge foundation and be able to take initiatives of what they want to learn.

      Comment by liyiwei — August 4, 2013 @ 8:42 am | Reply

RSS feed for comments on this post. TrackBack URI

Leave a Reply

Your email address will not be published. Required fields are marked *


This site uses Akismet to reduce spam. Learn how your comment data is processed.

Theme: Rubric. Get a free blog at