Ask HN: Pls. give me a hard prog. problem (re: talent wars)

https://news.ycombinator.com/item?id=2344284

by nadam • 15 years ago

2 2 15 years ago

news.ycombinator.com

This will be an unusual request for a job.

I've read Fred Wilson's blog post: 'The war for talent'. I am living in Eastern Europe (Hungary) and when I read posts like that (and similar Techrunch articles) I am always wondering: I know most people in the Valley are very talented. But are they that much more talented than people in other parts of the world? My yearly salary is $36.000. It is very good in my country. People in the valley earn $100.000 or more. Why are most SV companies not looking for help from other countries? Do they think communication problems are so serious that it is not worth it? Or do they somehow think that there is absolutely no talent in cheaper countries?

Here is a crazy request: give me a very hard programming (algorithmic) problem. Harder what you usually ask at interviews. (I usually very easily solve interview problems at typical companies.) Give me something so hard that if I can solve it in some days you hire me as a freelancer for remote work for $50/hour. (It is a crazy amount for me, but give me a very hard problem.) A problem can be for example something like this: Do this and this in O(nlogn) time and with O(logn) memory.

Does not matter if I will fail. I want to know: The reason I cannot earn SV-level salary is that I am not talented enough. Also I love thinking on very hard problems, I am addicted to them. (Once I was thinking about P vs. NP for a whole month. I learned tons of math during that.)

Please don't give me something which cannot be solved in 6-7 hours. I have a day job and two small children, so I can think only for 3 hours / day.

Thanks.

code• 15 years ago

Things are not so black and white. In my personal experience many times over, communication and ease of communicating correctly and how fast you move, especially in a startup environment, IS significantly better if not entirely necessary. I've had lots of personal issues making a long distance thing work even with last partners located in a different state much less a different country. I'm sure others can make it work but that's never been the case for me.

Second, in a startup culture early on, in my personal experience, we've had guys jam out all day and night, weekends, etc to push a product out when timing was critical. I can't honestly say I believe the same deadlines can be met with a remote team.

Finally, to be a great coder, beyond godlike programming abilities, you must have certain quality traits (too many to list here).

But per your request, here's a technical challenge we've done in the past. Find an insanely hard captcha, figure a machine learning algorithm for it to be able to read the captcha without relying on human eyes. As the mahine continues to view more and more captcha, it learns to decipher it. I will say you shouldn't do it for the purpose of breaking captchas and in some cases this is illegal but the technical problem to solve it is really good in terms of coding capabilities.

nadam• 15 years ago

"I've had lots of personal issues making a long distance thing work" Yes it is a problem. The best way it works is when you modularize your organization so that module boundaries are where the least communication is needed. This is how the business world works: The whole thing is modularized into companies, companies take ownership, specialize and communicat with each other mostly remotely. What most companies do wrong with offshoring is this:

- They don't try hard enough to find real talent. They are just searching for random people. - For the low amount of money they find random sub-par people. - They find out these people are capable only of munadane tasks. - They offshore only mundane tasks. - They think it is risky to give ownership to these people, so they try to 'integrate' them. They don't give ownership to remote teams.

In reality they could find talent and they could give ownership to talented remote teams. Do you know this?

http://netfixprize.com

The top teams contained insanely good quality people. By far not all of the top teams live in SV. (While it is well represented.)

"Finally, to be a great coder, beyond godlike programming abilities, you must have certain quality traits (too many to list here)."

This is also a prejudice of american companies (I am not saying you have this prejudice) that people in the 'east' can be good coders, but cannot be well-rounded. 'They cannot see the big picture'. This is not true. I was a math and algorithm enthusiast early-on (I started programming at the age of 12). I mentioned algorithmic problems in my post because this topic is quite objctive. But otherwise I am not a young coder. I am 36 years old. I am an experienced programmer. I know the importance of good software design. Managing big codebases and creating algorithms: these are different things. I have a bunch og cognitive patterns in both fields. I love linguistics, I love philosophy and psichology. I am interested in interaction design. I am interested in business strategies. I am a very intuitive person. On the side I am working on my startup which is about how ordinary/business people can intuitively create and access databases. (I am working on something like Excel for relational databases.)