Saturday, December 13, 2008

On Patenting

We in academia should be busting patents rather than defending them. Basically, the patenting process is stretched (with parties patenting things from silly to broad, gross, and to all things published and unpublished), with the courts and lawyers ultimately resolving the issues. No matter the issue, there are "expert" academics who will defend each of the opposing sides with equal vigor, authority and self-belief. Since we care about creativity and patents are a form of creative achievement, it should be our agenda to bust patents, and bring it the same level of angst we do for our papers (find flaws, missing elements, prior work, and with some small training, check even the legalese of claims, etc). This is doable with patents in algorithms.

ps: Imagine a graduate course on patented algorithms with HW to bust them, imagine a "algorithm zoo" of patents, and imagine a SODA submission on optimization problems in patent busting. :)

Labels:

2 Comments:

Blogger Hugh said...

but would this lead to fewer patents, or patents that are harder to "break" now that the community knows how to write a watertight application?

12:44 PM  
Anonymous Anonymous said...

Anything that brings more thoughtfulness (on the part of lawyers who draft them, researchers who supply them material and legal process that enforces them) to the patenting process is good, IMHO. Also, a different alignment of economic incentives (vs sweat) should help.
-- metoo.

4:36 PM  

Post a Comment

<< Home