y combinator thoughts

I wrote this on the plane in between reading "Pirates! In an adventure with Scientists". Here is hte blog entry in it's overly long, stream of consciousness entirety.

We had our meeting with Y Combinator yesterday. We didn't end up getting seed funding from them, and the reason that PG stated was surprising to me, but makes sense, particularly from an investor's viewpoint.

Paul Graham mostly said one thing during the meeting. It was not auspicious - he said, "Who would buy you?" Admittedly, a good question for a potential investor to think about. Not as important for myself - after all, I'd be happy with some good revenue that lets me earn a decent living. But an investor is going to want a liquidity event to profit off of their investment.

I think i did charm them with my various anecdotes of how I use CrikeyGuru to do everything these days, and my imagining of what one can do with the service. But they really wanted us to do it ALL ourselves. And I just see that as taking away too much development time from the other stuff we would like to build. I'm fine building that later if it becomes necessary - say mturk shuts down shop. We could build it quite quickly. But why reinvent the wheel? Why not rewrite PayPal while I am at it? Why not rewrite Apache to serve the web pages?

I used to be the kind of coder who wanted to reinvent the wheel, rewrite everything. But at some point I become lazier. I'm not at that "if it ain't broke don't fix it" point exactly, but I've gotten a lot closer to that mentality. I've learned to make the most of my coding time by using frameworks, third-party modules, and third-party web services like mturk. MTurk let us get up and running so much faster than if I had said, "Oh Mechanical Turk, what a great idea. Let's write our own!"

One of the reasons I love coding in Perl is all the available code out there. CPAN, the archive of Perl modules (like libraries) is Perl's killer app. Aaron Swartz wrote web.py for Python. I would never do that. It is already available for Perl and is brilliant. I don't have to write all sorts of things, because soemone else wrote it already. IT saves me a LOT of time. I've thought about coding in Ruby because Ruby is clean and easier to read and more object-oriented. But Ruby doesn't have WWW::Mechanize, which lets me easily use code to surf the web, which i use all the time to test web pages or check bank account balances. Sure, a community has to start somewhere, someone has to write the first code. But I don't have time for that, and I see no reason to leave Perl. I will do that when I see a that something I really need is no longer being met by Perl, or that some other language solves some huge problem.

My best guesses are the following: eBay - if we focus on customer service, answering all those questions that people on ebay email (you know, to ask about a product, follow-up, etc). Another large company like Yahoo! or Google, who are implementing their own mechanical turk like product. They could use our stores idea as a showcase product for their new product. But is anyone thinking of doing this?

Also, we met a guy who used to work at Amazon, on the Mechanical Turk project. He got YC funding for the winter session, and was hacking in the office that day. He couldn't tell us much as it was obvious he was still under NDA with them. But it was interesting to talk to him some and work out some puzzle pieces.

Trevor Blackwell seemed quite interested in our idea, Robert Morris didn't say a single word, Jessica Livingston desperately needs our transcription services for her hours of audio tape, and Paul Graham mostly wanted us to reinvent MTurk.

I was joking last night that the problem is Nathan and I are lazy - we don't want to do everything ourselves. Hell, I outsource just about everything I can to MTurk these days - I outsourced customer service the other night, I outsource phone calls I need to make. My latest idea is to get some sort of organizational method that uses MTurk to figure out what the hell I need to do and prioritize my day for me.

Hmm i should wrap up this entry now. basically, i think we are better off working with Amazon then going it alone. They are offering us tons of great help - in many ways, help a lot like what YC would offer. But all free.