4 ms·
To speak to this directly in a scientific setting I find one of the most beneficial thing I can do when exploring a new domain is to NOT read works first. But i
by godelski 3y ago
To speak to this directly in a scientific setting I find one of the most beneficial thing I can do when exploring a new domain is to NOT read works first. But instead to first ponder how I would go about solving the problem, maybe do some naive attempts, and THEN read. An iterative cycle of this becomes successful because of a few typical outcomes: you spend more time but generally come away with a deep understanding of motivation and why the field has moved in the direction that it now exists in (which helps you understand where to go next!) and/or in this process you find assumptions made to simplify a problem but has been forgotten and can be readdressed (at worse you have a deeper appreciation for the naive assumption). I do think this process is slower, but I find it frequently results in deeper understanding. It is (for me) faster if your intent is to understand, but slower if your intent is to do things quickly. I find that with the typical method of review to catchup without the struggling generally results in me not having a good understanding of many of the underlying assumptions or nuance at play (I think this is a common outcome, beyond my own experience). YMMV and it may vary dependent on your goals with a specific area and cross-domain knowledge. But I find great pleasure in it because it causes me to consider many things beautiful and genius where I would have otherwise saw them as obvious and and mundane. That alone is enough for me because research is a grueling task with endless and frequent failure but this helps keep me motivated and keeps things "fun" (in the way I think Feynman was describing). It's not dissimilar from doing your homework, checking someone else's work/a solution manual, and then trying to figure out where your mistakes are rather than simply correcting them. Similarly, time constraints are a pain so such a process isn't always possible and certainly not as often as I'd like.
- mistermann 3y agoSounds were quite familiar to this anecdote from Feynman about his learning path: https://literature.stackexchange.com/questions/8691/in-what-book-does-paul-dirac-say-some-new-ideas-are-here-needed https://literature.stackexchange.com/questions/8691/in-what-...
- godelski 3y agoYeah I've actually learned a lot from Feynman about how to learn. I think a lot just comes down to acting like a kid. Fuck around and find out, but with more nuance and sometimes more direction or not haha. There's the old saying "passion is worth 10 IQ points" and I think these all are part of the same thing.