Fun with Presentations!

A brief behind-the-scenes overview of the presentation on the paper “How to Play Any Mental Game” (by Goldreich, Micali, and Wigderson) I gave today.

[Fun fact: the paper is so old (from 1987), the figures are hand drawn.]

Weeks ago: I need to give one more presentation. Let’s see what papers haven’t been taken yet. Hey, there’s one called “How to Play Any Mental Game.” That sounds interesting.

Wednesday: I better start reading this paper. [time passes] I’m screwed.

Friday: I’m screwed but I have to do this anyway. Better read it again to see if I can understand it now. [time passes] Wait a minute, it says if Alice has bit a and Bob has bit b, Alice can compute a AND b without learning b. That’s bogus; the result gives away the value of b! Did I just debunk this paper?

Saturday: No! When you compose ANDs, Alice doesn’t learn the intermediate results, so that isn’t really a problem! This is actually starting to make a little sense. Better start writing up some slides.

Sunday afternoon: Time to practice. [time passes] Yikes. So many errors in the slides. So much fumbling.

Sunday night: Time to practice again. [time passes] OK, a little better. Again? [time passes] I’m ready.

Monday: Here we go. [kicks ass, takes names]

3 Responses

  1. I miss the days when I could freely connect to ACM to download the full-text from my dorm. I’ll check to see if work has access.

  2. If you can’t get your hands on a copy of the paper, I could probably help you out. Old papers like that tend to be hard to find.

    The one line summary: if you have secure private channels and synchronous communications, it’s possible to do secure multiparty computation of an arbitrary function as long as less than half of the participants are malicious.

  3. What is this? Paul procrasinated? Apparently Math Blaster didn’t do too good of a job teaching you time management :) Glad you got the presentation done.

Comments are closed.