Advice to a Beginning Graduate Student (2001)
2 hours ago
- Treat books as random access; don't read from cover to cover; skip and find understandable parts.
- Write in margins and write while reading to better understand hard material.
- Writing gives you the power of a Turing machine; without it, you're like a finite automaton.
- When stuck, give yourself a hint like 'I could tell you something' to find shortcuts.
- Learn from solutions you couldn't find or that others published; ask how you should have thought.
- Embrace paradoxes and contradictions as powerful sources of insight.
- Try small examples and put yourself in the middle of a solution to approach problems.
- Don't expect your advisor to give you a solvable problem; work hard and communicate progress.
- Focus narrowly on a topic to eventually see the world in a grain of sand.
- Try to disprove your own beliefs; it gives insight even if they turn out true.
- Be open to answers that are not what you expected, like randomized algorithms for NP problems.
- Have something to say, say it, stop, and give an accurate title.
- Have peers read your thesis for feedback; don't rely solely on advisor.
- Use rejection feedback to improve your paper and resubmit.
- Aim to be unique and special in some way; choose an inspiring end and do it.