I was really pleased when Terry started blogging

...rarely does reading a blog push me past my current range. smile

There's a promising trend of putting, on blogs or wikis (Terry has done an incredible job on the dispersive wiki, and Scott Aaronson built the complexity zoo), thoughts not quite ready for 'primetime' (nor any specific journal). It's quite an exciting development. It certainly changes the social dynamic: I can imagine that in the near future, the most important research institutions will not be individual research departments but online communities crystallizing around prominent scientists who happen to be terrific bloggers.

I think the way to translate this principle for hackers is: "Continually aim to be the stupidest person in the room".

If you want to improve quickly you must work on projects with people who have a lot more skill and experience than you do.

As long as you have sufficient ego to not take the experience of being repeatedly outperformed to mean that you suck.

Most programmers are not nearly as good as they think they are. Learning how much you suck at programming is the very first step to getting any good at it.

Oh, please, can anybody correct the title? It's "Fields medalist", not "Field's medalist".

For many years, weight lifters thought that by increasing the number of repetitions, they could increase their strength. Eventually, they found out this wasn't true. What's needed is a very few reps and a great amount of weight. The same is likely true with intellectual pursuits. Better to struggle with Bach's inventions 30 minutes a day than to play "Smoke on the water" for 10 years.

Very interesting! Do you have some source / link for this?

You would probably want to go for size, not just strength, so about 10 reps per set instead of 4 or 5, about 3 sets plus a warmup set, and about 3-5 exercises per body area, and repeat about once a week. So, about four times per week. You really want to take about 45 to 55 minutes, unless you do cardio for 20 minutes beforehand to build up cardiovascular strength as well, in which case you would spend about 1:30 at a gym, four times per week. That's all it takes. Tomorrow, I'll try to find some links.

But the main thing is to eat a lot of protein, drink plenty of water (to flush your kidneys), eat many meals throughout the day, and get lots of sleep. That's what's harder to do than going to the gym.

The original author is actually wrong..especially using his chess analogy. Incremental improvements will get you nowhere. The best way to raise your chess game is to play against people who are faaar better than you. Do this for one week, and you will find yourself crushing your [former] peers.

Using the same analogy, Mark Zuckerberg should have beeen aiming to build facebook into a 10 million dollar company (or something)

That point about chess is highly debatable.

Brain scans on domain-specific experts (math and chess are domain-specific skills) have shown activation in areas that store episodic memory. What this means, is that a large part of the "expertness" comes from efficient retrieval from long term memory (a fast pipeline between RAM and HDD, if you will). Writing to HDD, in the brain at least, is usually an incremental process.

I am not a chess player, but in go, this makes perfect sense. You start with the most basic strategies and build on them as you advance. If the skill gap is too wide you'd probably drown in confusion trying to take apart what is going on.

Granted, it's not black and white. If you're smart you'd be able to take bigger jumps and piece together several steps with each jump, in which case, for you, playing somebody faaar better is simply a reasonable next step for you, but relative to your skill level it's still a "slow" process.

The original author is arguably the world's best mathematician. He certainly hasn't gone nowhere. It may be worth considering that he knows what he's talking about.

It may be worth considering that I am talking about chess.

You made an analogy to Facebook, so you weren't just talking about chess.

I will have to disagree with you there. If the gap between you and your coworker/chessopponent is too great there are no lessons learned. Trying to assimilate knowledge too fast will only lead to superficial understanding and frustration.

Sadly, you're describing me and quantum mechanics.

I think it doesn't matter if you play people who are slightly better or far better players than you are. The secret is getting your ass kicked as often as possible.

Bach's inventions aren't really that hard.

...they're used in education, unlike e.g. Rachmaninoff's 3rd.