One Question at a Talk Can Be a Return Offer

On a stormy Sunday afternoon in 1988, Edmund Clarke, a Carnegie Mellon professor and Turing Award winner, came to lecture at Fudan University and found the room nearly empty.

Author: Koutian Wu; GitHub: ktwu01

Lu Qi was pulled in at the last minute to fill a seat. During the talk, Lu Qi asked questions that left a deep impression on Clarke, who was clearly interested in this slight young student and even looked at his research papers. In the end Lu Qi was invited to do a PhD in computer science at Carnegie Mellon on a full scholarship. Even the 45 dollar application fee was waived.

The story sounds almost mystical: Lu Qi just happened to be dragged in to fill a seat. But in reality it could not have been improvised on the spot. There must have been deep preparation behind it.

So whatever else happens, you need a foundation first.

Then comes the emergency. With a foundation, even in an emergency you can still stand out, because you can pour a lot of effort into a very short window. Say the talk is in the afternoon. Then skip every class you have that morning, don’t attend any of them, and spend the whole morning preparing for this one thing. I think that is very likely what Lu Qi did.

Why do I think so? Because I have seen it myself.

A few years ago Shing-Tung Yau came to give a talk at USTC. We happened to have some unimportant classes that day, and most students really did skip all of them to hear Yau. In the end most of us didn’t understand the talk, but some people really did ask good questions.

Another time, Chao-Yang Lu from Jian-Wei Pan’s group came to give a talk. One student who was very interested in quantum information asked him several questions afterward that Professor Lu could not answer. Professor Lu was delighted. After the questions, he assumed the student was at PhD level, and was surprised to learn he was only an undergraduate. He invited him to join the group and write papers. A few years later, the student had a paper in PRL, Physical Review Letters, one of the top journals in physics.

Lu Qi and my schoolmate set a very good example for the rest of us.

Is skipping class to attend a talk worth it? Do the math. Most of what is taught in your regular classes you could learn from the textbook. But how many times a year does a Turing Award winner or the head of a top lab come to your school? If you can make that person remember you, the payoff might be a spot in the group, a recommendation letter, a return offer. The classes you skip are a linear loss. The one chance you catch can be an exponential return.

The precondition is that you actually prepare. Skipping class to sit in a talk scrolling your phone is pure waste. Faking understanding and asking an outsider’s question leaves a worse impression than asking nothing.

So what do you do with that morning? Read the speaker’s homepage and recent papers, enough to know his direction and his latest results. Prepare three to five questions that a quick search can’t answer and that the slides won’t cover, ideally ones you are genuinely curious about, such as where his method breaks down or whether it extends to other problems. Listen carefully during the talk and adjust if your questions get answered. When you ask, open with a line like, I’ve been reading your work on such-and-such. And don’t leave right after; if you get the chance, go up and keep talking.

People ask how they know they are ready. My test is simple. Can you explain to someone, in three minutes, what this person works on, what his core contribution is, and what question you have about his work?

Chances like this won’t come often in an academic career, maybe a few times a year.

Build the foundation day to day, and when the chance comes, have the nerve to go all in. You need both.