Ask HN: Does anybody else feel overwhelmed while reading HN?

Absolutely. It's much worse if, instead of HN, you follow a niche area like machine learning because the pace of progress is so fast plus each paper / project that gets released is so dense. It took me a couple of weekends just to set up an old box with Linux and the proper drivers for a GPU, learn python virtualenvs, etc; Meanwhile, it's absolutely discouraging to look at reddit.com/r/machinelearning and see the flurry of productive activity.

I think a sense of resignation is actually useful here. Just resign yourself to the fact that you'll never be as good as them and that it will take you 10 years to be able to just follow instructions under a Google or Facebook AI scientist (, say). And continue to trod on like the tortoise in the tortoise vs. hare story :-)

I see nothing in this comment that would require a downvote.

You might appreciate Ira Glass's take on this. https://www.youtube.com/watch?v=X2wLP0izeJE

I've been a professional web developer for 18 years now. I was very much on the bleeding edge of web standards, and jumping on Rails in 2005, I did everything from managing my own servers through backend, front-end and design in Photoshop. At the time the web was still a very greenfield type of place that was uncolonized by the top talent in either software engineering or in design, as a result it was possible to be one of the best web generalists with a little bit of aptitude and a lot of interest.

Fast forward ten years and every discipline of web development now goes very deep. It's still worth it to have a broad skillset, but it's no longer practical to be upper echelon across the board in web development. This generally leads to a feeling of overwhelm and regret that I can't learn all the things I possibly might want to learn, but on the bright side the playground is bigger than ever.

My advice is don't spend too much time thinking about the big picture, instead pick one practical project at a time and spend 95% of your time making it the best you can. Even if you only read HN a couple times a month, that's all you need for basic awareness of the landscape. By giving yourself heads-down time you can replace some of the overwhelm with a feeling of accomplishment, and you'll be growing your skills to boot.

Oh, you will feel much better once you have mastered:

1. Programming chips in binary, machine code, and C. You need a variety of chips. Try to learn at least 5 from each manufacturer.

2. Learn at least 37 Javascript frameworks, as evenly divided between front-end and server side as you can. (Good news: angular 1 and 2 count as 2 frameworks).

3. Learn Scala, Rust, Haskell, C, C#, Java. (Python and Ruby go without saying).

4. Learn R, machine learning, statistics (prob and regressions), linear algebra and multi-variate calculus.

5. Learn growth hacking (edit:) and lean startup, human centered design, and design thinking.

6. Learn accounting, finance (go through Markowitz, to Black Scholes, Fama, CAPM, and factor models. Read the original papers only and implement everything yourself, in 2 languages).

Now you are ready to read HN.