[PRL] 5 minutes

Matthias Felleisen matthias at ccs.neu.edu
Wed Feb 8 18:38:29 EST 2006


Now that you have a few hours to think about your 5 minutes, let me 
suggest you rethink it with the following constraints in mind:

1. Assume little about the "target"'s knowledge or level of 
sophistication.

2. Assume the "target" is a computer scientist. Do not assume he can 
program. By no means, assume he can program or has experience 
maintaining programs. (Except for tex programs and he's bad a that).

3. So tell the story, starting from Adam and Eve. From there, describe 
the problem. Get there in 2 minutes.

4. Then tell him that "nobody" has looked at this problem from "your" 
angle and what that is.

5. Then explain what that is. These last 2 1/2 minutes are your story.

;; ---

If you are working on a long-standing problem, you can mention names of 
people who have worked on this. For example: "I am working on the 
full-abstraction problem for PCF, picking up where Milner left off in 
1972. I am using Curien's trick of SDFs but borrow game semantics 
terminology from the Abramsky school."

Don't say "my co-authors are famous and therefore I am working on good 
stuff." That doesn't work.

-- Matthias




More information about the PRL mailing list