Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Learning to Learn

kevin.the.li

Post reply on HN
Print view View original

31–40 of 142 posts

Re: Learning to Learn

#31

Post by server_man3000 »

cbracketdash wrote: post #18
keeptrying wrote: post #15

A big hole in this article is that you need to find the very best learning resource there is. This is a must. Eg: For RL it would be Barto&Sutton book. Sometimes the best source is not intuitive. Eg: The best way to become a safe driver is to go to performance drivign school - its a bit expensive but they tell you how to sit and stay alert in a car which I have never seen outside of these schools. One of my most comm…

Would love to hear more about your thoughts on learning medicine! Why would you recommend an online program over textbooks?

When I’ve gotten deep into a topic I’ve actually almost ALWAYS learned that textbooks are the best way to learn things.

The internet is full of information. Sometimes it’s too much, unstructured or tangential to the goal at hand. Textbooks, in my experience, are truly written by the experts. It’s been vetted, rigorously reviewed and fact checked. It’s not inspired by influencers or clickbait.

Obviously YMMV, but when you find a top recommended textbook, it’s usually miles beyond a YouTube video or medium blog for deeper level content. It usually flows better and makes more sense as you study consistently.

Re: Learning to Learn

#32

Post by sam29681749 »

keeptrying wrote: post #15

A big hole in this article is that you need to find the very best learning resource there is. This is a must. Eg: For RL it would be Barto&Sutton book. Sometimes the best source is not intuitive. Eg: The best way to become a safe driver is to go to performance drivign school - its a bit expensive but they tell you how to sit and stay alert in a car which I have never seen outside of these schools. One of my most comm…

It's not foolproof, but some universities publish their course textbook lists online (and in some cases recommended readings too). As a bonus, textbooks often have recommendations for further readings.

Re: Learning to Learn

#33

Post by dartharva »

eacapeisfutuile wrote: post #27
dartharva wrote: post #23

Earlier quoted context omitted.

I checked these out but couldn't find the answer to the questions I am asking. It tells you how to maximize efficiency while learning, but doesn't tell you how to identify what you should be learning in the first place and what the sequence of your focus should be when you start.

What do you specifically want to learn? What you need to learn will generally make itself known, if you have some general goal? Just starting at all is the best way to start.

Say I want to learn web scraping. Now if I try to start, I see there are several layers of fundamentals and several potential paths, all of which seem equally important. Where do I start? What platform should I use? Should I start from the DOM and HTML parsing and go ahead from there, or should I start from learning Python libraries built for it as so many books and tutorials advise? Going one layer deeper, do I need to learn data structures before anything else? JavaScript and APIs? The fundamentals of TCP networking? Regex?

It's all too overwhelming at times.

Re: Learning to Learn

#34

Post by eacapeisfutuile »

dartharva wrote: post #33
eacapeisfutuile wrote: post #27

Earlier quoted context omitted.

What do you specifically want to learn? What you need to learn will generally make itself known, if you have some general goal? Just starting at all is the best way to start.

Say I want to learn web scraping. Now if I try to start, I see there are several layers of fundamentals and several potential paths, all of which seem equally important. Where do I start? What platform should I use? Should I start from the DOM and HTML parsing and go ahead from there, or should I start from learning Python libraries built for it as so many books and tutorials advise? Going one layer deeper, do I need…

Yeah, start from any point A. A suggestion is where you can progress in some way. Don’t make too many decisions. There’s so many viable ways to learn that, just start where it is natural for you. You don’t have to figure everything out or have answers to everything.

Re: Learning to Learn

#35

Post by dartharva »

eacapeisfutuile wrote: post #20
dartharva wrote: post #2

This is something I have personally struggled with, so I wish the author elaborated more. If you are a novice, how do you quickly identify what the foundational knowledge is? How do you know what makes you an expert and not an "expert beginner" as the author says to the extent that you can build a personal curriculum about it?

You don’t. That is what you will build as you learn a topic, it makes no sense to “identify” it first. Do the thing you want to learn.

I learnt data analytics and SQL for MIS/BI purposes on-the-job as a sales manager. Got pretty good at it too, built several dashboards and long-standing capabilities for my team.

Now, if I say I want to get into this scene for good, I am immediately daunted with a mountain of diverging learning paths to take. Should I take to Python and its massive library ecosystem, or should focus on database fundamentals? In every choice taken there are seemingly infinite branches, and it is rather hard to focus if you aren't even sure you're on the right track.

Last time I sat in an analytics/consulting interview they grilled me on highly specific technical questions on data pipelines and warehousing and testing and other topics that I've never had to worry about before at work. In another assessment, I was grilled on some AWS/Redshift-specific things. In yet another I was expected to know deep learning. It is all too hard for someone not originally with an engineering (or adjacent) background.

Re: Learning to Learn

#36

Post by esperent »

zfnmxt wrote: post #26

Posts like this that talk about learning "efficiency" always come off as soulless and dystopian to me. I think learning should be fun and that fun learning is the most effective---that's the only thing I optimize for and I certainly don't think about efficiency percentages. What a drag that would be.

I studied mathematics and I think this subject illustrates my feelings on this perfectly. Coding, just as well.

When it comes to learning maths, or a new programming language, there's all this tedious boilerplate you need to know. The rules, or syntax, the names of everything, how it all fits together.

There's ways to make learning this stuff more fun, but ultimately, not that much more fun. And anyway, the learning part is not the good part, it's the things you can do once you reach a certain knowledge level that are incredible, beautiful, even sublime.

On the other hand, take something like learning to paint, or taking dancing lessons. Unless you're hoping to become a member of an international ballet company, learning to dance is the fun part.

As another point, if you're a knowledge worker and you're likely to have situations in your life where someone basically says to you "right mate, you've got the job, here's a huge body of deep technical knowledge to learn, get up to speed, see you Monday" then a certain amount of skill in knowing how to absorb that quickly is a good thing.

Re: Learning to Learn

#38

Post by eacapeisfutuile »

dartharva wrote: post #35
eacapeisfutuile wrote: post #20

Earlier quoted context omitted.

You don’t. That is what you will build as you learn a topic, it makes no sense to “identify” it first. Do the thing you want to learn.

I learnt data analytics and SQL for MIS/BI purposes on-the-job as a sales manager. Got pretty good at it too, built several dashboards and long-standing capabilities for my team. Now, if I say I want to get into this scene for good, I am immediately daunted with a mountain of diverging learning paths to take. Should I take to Python and its massive library ecosystem, or should focus on database fundamentals? In every…

Yeah learning to actually use in practice is different from preparing for interviews. I would say continue where you already built some knowledge and branch out when you realize you have to, and then work on learning one branch as you go.

For interviews you may need to lookup what type of questions to expect and memorize details on that, unfortunately. That is not useful in practice but can be necessary for interviews.

Re: Learning to Learn

#40

Post by blackoil »

zfnmxt wrote: post #26

Posts like this that talk about learning "efficiency" always come off as soulless and dystopian to me. I think learning should be fun and that fun learning is the most effective---that's the only thing I optimize for and I certainly don't think about efficiency percentages. What a drag that would be.

I would say "fun" is overrated. We have become so focused on everything being fun that everything including fun itself has become tepid and mediocre. It is important to slog through the hard parts to cross the barrier of expert beginner. We are over downplaying the value of hard work and grit.
Post reply on HN