> In order to achieve consistent deep work, I usually need to first immerse myself in the relevant research literature, seeking a problem that it is recognized

> In order to achieve consistent deep work, I usually need to first immerse myself in the relevant research literature, seeking a problem that it is recognized as important, unanswered, but probably answerable with my skill set. This is not easy. I want to run out and repeat this to every PhD student I've met in the last 2 years.