It's the future. Crowdsourcing through job candidates.
I had a similar experience with a company in Berlin, Germany. They didn't care about the code or my skills. Only if the submitted solution worked against their existing tests.
I think it's just a sleazy way of getting things built for free. Basically, write only a test for the functionality you want. And give the actual task to job candidates. If someone succeeds, you have both the algorithm and its test for pennies. It's based on P = NP. Testing the result is easy. Designing a solution is the hard part.
I now refuse both technical interviews and take at home tests. Limits my choices a lot. But the ones that remain are usually top shelf.
Wait, you refuse technical interviews and you say the rest are top shelf? I'm curious to know what your ideal interview looks like. Or do you mean you just get hired through networking?
I've had a very similar experience to this commenter. I wouldn't say that I reject all technical interviews and take home tests, but I certainly reject a lot of them and find that I have little trouble finding really high quality companies that will speak with me. As you say, a lot of it comes through broad and deep networking. I know people, I have a proven track record in business and technology, and people are comfortable working with me.
An ideal interview is one where I get to learn as much about who I will be working with and what I will be doing as they will be learning about me. Generally these turn into more discussion based interviews where we talk about the issues the company is having and how I might be able to best solve them. They still get a good idea about how I think and work by discussing the problem back and forth. I also get to see how they think and work.
Frankly, I white board problems very well. But when companies ask me to do so in an interview I want to see them do it as well. I want to see their ability to code as much as they want to see mine. Same goes for a take home test. If we are going to work together and you have to see code of mine, I very much want to see yours to make sure it is up to my standards.
You would stand a good chance of waiting a lot longer to get that work done, but as mrcold said, testing a solution doesn't cost much. If you're willing to wait for the solution to a problem, getting work done through code exercises can be worth it.
Yea, I guess I was thinking more that it would take a lot of time to unnaturally break down a problem into portions small enough to pass as code exercises.
I had a similar experience with a company in Berlin, Germany. They didn't care about the code or my skills. Only if the submitted solution worked against their existing tests.
I think it's just a sleazy way of getting things built for free. Basically, write only a test for the functionality you want. And give the actual task to job candidates. If someone succeeds, you have both the algorithm and its test for pennies. It's based on P = NP. Testing the result is easy. Designing a solution is the hard part.
I now refuse both technical interviews and take at home tests. Limits my choices a lot. But the ones that remain are usually top shelf.