I can't say that I've quit my job, but Paul Graham does make some interesting points.
The question that is stuck in my head is "what should I do if I have an idea for a Web service, but have no technical 'know-how?'" Is it worth approaching YCombinator?
Start ups are about work, not ideas. Ideas flow freely with some people (like me) that just have about one idea per day worth implementation.
The skill to bring a product to life is the critical part.
Still, since you shouldn't drop the idea of starting a start up, you should either find someone with the technical know how (track down my email and email me for help in that) or go for an idea where you can contribute 100%.
Just my opinion, of course, though I highly doubt anyone would disagree (yes, this is a challenge).
I remember reading a chapter of Hackers & Painters in which PG states that he was able to implement new features so quickly, that it seemed to his competitors as though he possessed some kind of secret weapon. Instead, he was simply using Lisp, a powerful and highly abstract language.
What he doesn't explicitly mention is that he is an expert at Lisp, having published two books on the language -- not to mention having a PhD from Harvard in Computer Science, where he probably became an expert using the language (although, admittedly, I have never been able to track down PGs graduate research work...).
This is what gave ViaWeb their competitive edge in the section of the e-commerce market and led to their success. As a startup, you're unlikely to succeed. You'll want all the advantages you can possibly attain.
Aur, thank you for your insight. Currently I would say I fall more into a marketing/sales/negotiation "box," which I believe is an important aspect for start-ups.
If I were to begin a start-up it would be as a co-founder. Yes, I would need to find someone with the technical know how - fortunately I'm living with a software engineer. Unfortunately, I'm pretty sure he is happy with his job...but maybe I can convince him to give me 20% of his time if I develop my plan further.
no technical know how when building a web company is going to be a problem.
But I'm building a web startup, and I'm not an engineer/developer. I do know a lot about the web and product development so my eyes don't roll back when the conversation turns to apache logs, SSL stamps, and server side scripts.
I would suggest documenting your concept in a simple word doc, starting with the primary result, then spell it out step by step.
After you have the core concept in place, talk to web developers about the project, if the idea is hot, and they have the right skills you could be on your way to success.
I have rather the opposite problem. I'm like a machine when I get a decent idea (wrt implementing it etc), but I can't seem to find one that makes other people care about it.
The question that is stuck in my head is "what should I do if I have an idea for a Web service, but have no technical 'know-how?'" Is it worth approaching YCombinator?