Pretty close to all the gold medalists at last year's ACM competition were regular competitors on TopCoder, including the majority of the MIT team (and the coach).
The role of culture can't be underestimated: in some places academic competitions are a really big deal. There are major centers for it: St. Peterburg in Russia, Waterloo in Canada, Zhejiang University and Shanghai Jiao Tong in China, and especially Warsaw University in Poland.
Just competing in TopCoder over the past year or so, I've solved literally hundreds of tricky algorithmic problems, under heavy time pressure, exactly. The best competitors have solved literally thousands of difficult challenges. It's really no wonder that they do so well. However, I don't think it's unfathomable discipline: I think it's an environment that encourages it.
I've never been any good at programming competitions, so to caress my ego I should join the crowd claiming that the results are meaningless. However, my experience is that people who perform well in programming contests tend also to do extremely well as software engineers. Perhaps I've just seen not big enough sample, but I don't think so.
I'm not saying people who do well in programming contests aren't smart, just that there are also smart people who don't do well in programming contests, e.g. because they don't spend time practicing the weird sort of programming you need to be able to do.
In order for your observation to have weight, you would really need a number of examples of people who couldn't cut it in a competition doing poorly as software engineers.
I always wonder why the people who write this article fail to compare programming salaries to other professions. The writer gets all confused that the best and brightest young Americans aren't chomping at the bit to get into a field that pays "top programmers upwards of 100K a year" (paraphrasing here).
Really, a whole 100K? Is that really so great, when you consider the options that the best and brightest young Americans have? A law grad from a top school with no experience whatsoever typically earns over $150K in his first job out.
I'm not here to complain - there are all kinds of great reasons to be involved in technology and programming. If I wanted to be a lawyer, I'd go to law school. There's huge money to be made in software, especially in startups.
But c'mon. Do you think the writers of these articles (because they happen over and over) just haven't thought of this? They do this every time, and never seem to consider the context (ie., other career options). Is it a deliberate omission?
another irritating one is the way they quote 900K engineers out of China and India without considering the way the US/India/China define "engineer". I did math as an undergrad followed by an MS in engineering, and I don't think I'd be counted (certainly if I'd done a grad degree in a math department - studying the same material, I wouldn't be counted). Meanwhile, it sounds like a 3-year trade degree in China would count as an engineer.
Again, there are reasonable interpretations of these numbers - I'm not saying the writer has to take my point of view. But to just chuck them out there as fact without any discussion is at best lazy journalism (again, it's such a loud omission that I start to wonder if it is deliberate).
Engineering is one of the better defined/controlled degrees in India. Any accredited engineering degree is a 4 year degree and the core curriculum includes as much Math as a US engineering degree.
However, very few of these schools and professors are anywhere close to the good schools and professors in the US.
I'm not at all surprised to hear that there are organizations in India that take a more restrictive definition of engineer. The problem, I think, is that we're comparing a lenient count in India and China with a very restrictive count in the US. When you count more similarly, the gap vanishes - and actually looks more favorable for the US on a per-capita basis.
The new courses that are designed to apply programming to real-world problems seem really promising.
In my opinion, giving up concerts, outdoor parties, and sunbathing - in order to have students practice programming for 3 hours a day so they can compete in a contest - would be a real shame.
Talk about a basic lack of perspective! Which is better: the student who plays competitive tennis, studies music, art and literature, in addition to competing with the Mathletes; or the student who studies Mathlete problems as her only hobby?
That depends an awful lot on how you define 'better'.
The guy who practices mercilessly and with focus on a single task is likely to excel at that task. Will they be a strong athlete, a fountain of conversation, a handsome deal-maker? Perhaps less likely. But if you want to get good at something then you practice that something.
There are only 24 hours in a day. Use them well...
Discipline is routinely mistaken for intelligence, because it does all the things intelligence claims to do (as opposed to what it actually does, which is significantly more limited).
Intelligence is an abstract measure, hard to measure and it's utility is unclear. Discipline, on the other hand, pays rich returns. Just in the last week there have been two articles on news.yc about how "experts" in every field are much more likely to be people who get average scores on various intelligence metrics like IQ but worked harder and longer on the same problems.
If you would like a more authoritative source that backs up this claim with extensive research take a look at the Cambridge Handbook of Expertise.
Isn't intelligence known more for ground breaking work than becoming an expert in a particular field?
Also, I think a big reason we don't see more results from intelligent people is due to our cult of intelligence that worships the raw power more than people making good use of it.
But now India is at a turning point. It's getting to be a leader
A leader at what? I can't offhand think of any software that I use that comes from India or China. I can think off a couple of things from Eastern Europe. Not much.
Programming contests? Eh. Programming by itself is completely worthless. It has to be matched with some problem...and such problems are not found in front of a monitor.
The only thing India is a leader at right now is developing outsourced software and services. I agree we need to see more finished products coming out of there. One of the reasons this doesn't happen is (I think) you can't develop a product while sitting thousands of miles away from your customers. You need a local market for whom you make the product and then if it's good you can export it. The local market in India for most types of software is still small (but growing).
China supposedly has top notch programming going on for its military. Second (only?) to the US. Otherwise, they're far behind. India... nothing to brag about.
The role of culture can't be underestimated: in some places academic competitions are a really big deal. There are major centers for it: St. Peterburg in Russia, Waterloo in Canada, Zhejiang University and Shanghai Jiao Tong in China, and especially Warsaw University in Poland.
Just competing in TopCoder over the past year or so, I've solved literally hundreds of tricky algorithmic problems, under heavy time pressure, exactly. The best competitors have solved literally thousands of difficult challenges. It's really no wonder that they do so well. However, I don't think it's unfathomable discipline: I think it's an environment that encourages it.