No Defense

I was at a PL / Compiler retreat last weekend where the format of the talks and topics were presented in the standard fashion. Everyone talks for 20 or 30 minutes with a few minutes at the end for questions. However, the conference asked for something quite unique and actually brilliant. A microphone went around and everyone who did not present or was not part of the group was asked for their feedback in front of everyone. The people who presented were forced to listen, they could not defend their work. They had to listen.

What an amazing yet frightening experience for all the conference presenters. You have just presented your years of toil to be potentially torn apart in front of you by your colleagues. For most people, including myself, your first instinct is to defend your work and tell the reviewer how they are wrong. It's a bad habit, but it happens. That's why, most of the feedback we get is usually in a softer tone, through a note, an edited paper, or as a soft indirection - "maybe you should check this out." The idea that you have to sit down, shutup and listen is an amazing feedback mechanism and I applaud the conference organizers for doing it.