Grad School Advice

If I learned anything from grad school, it is that you need to think in the audiences' role.

Paper: reviewer
Presentation: listener
Company: customer

换位思考，知人所想。

Acceptance. Appreciation. Respect must build credit first to win people's trust.

Rules:

Always go for extra miles. 10% more, 20% more
No matter what you do, go for your goals, go deep and zoom in.
Have examples, easy to remember, easy to explain

Research

Answer this question first!

What is unique about your paper, why others' paper does not work? What's fundamentally different? What is blocking others from doing this?

You need to be able to summarize the key points to others in 3 minutes.

Presentation

WHO is the audience? Introduce the aspects that your work interests them! What they are interested.
Start with high level points (using something that users are familiar with, but not tedious details). -what is the essential part of the problem? Why they are important? What are the challenges?
Use more data from Google/Facebook, do not just use paper citations. 1'000 papers does not matter.
How to design talks Key takeaways are to use good story with examples figures in design slides. Always provide enough context clearly before delivering ideas.

Writing good emails

Thanks in the front or say sth nice, before you jump into the things that you want others to do.
Thanks again in the end.

Work with advisor

When idea gives you some idea, you do not have a sense of urgency, you stuck somewhere and go back to your advisor, say hi your idea does not work. This is not how it works. Advisors also have pressure from funding agency - they also need to deliver.

Research advisor is not like your boss. If it does not work out, an active student can change it yourself by talking to others. Waiting for advisor to give suggestions is not the only solution.

Most of the unsuccessful students I've observed fail not because they're not smart or hardworking, but because they get stuck for extended periods of time and grow demoralized. Keep moving.

A healthy model between advisor and students are that students works for their own project, provides enough info for advisor to discuss in the meetings. This is the good way to get suggestions. Or if really stuck, show what is the problem, what have been tried, what are the results of your experiments. If you do not have these information, no one can help you.

The good point to leave grad school is when you feel you can not learn much from there. Some empirical observations for graduation standards.

Time management/Work habit

How much real work time do you have each day? Analyze and optimize.

Work attitude

The Genius Fallacy by Jean Yang

What I have learned is that discipline and the ability to persevere are equally, if not more, important to success than being able to look like a smart person in meetings. All of the superstars I've known have worked harder--and often faced more obstacles, in part due to the high volume of work--than other people, despite how much it might look like they are flying from one brilliant result to another from the outside. Because of this, I now want students who accept that life is hard and that they are going to fail. I want students who accept that sometimes work is going to feel like it's going to nowhere, to the point that they wish they were catastrophically failing instead because then at least something would be happening. While confidence might signal resilience and a formidable intellect might decrease the number of obstacles, the main differentiator between a star and simply a smart person is the ability to keep showing up when things do not go well.

Funding agency

NSF is slightly better, one year review.
DARPA requires monthly meeting, poster, three-month PI meeting, final review. DARPA is good, but the requirements are mad.
Industry funding still need some publications to show academic collaboration results.