This is an excellent overview.

* His points about collaboration are excellent. So many research students think that their brains are their best asset. There are many smart people. There are much fewer smart people who can communicate and collaborate well. Be one of those people.

* His points about papers are completely on spot. There are simply too many papers, and many actually aren't that good. The light bulb moment for me was realizing that this author didn't write this paper to help me solve my problem, they wrote it to describe how they solved their problem. Finding the right way to measure whether "their problem ≅ your problem" is key.

* Do a lot of learning on your own. If you don't you're only hurting yourself.