Learning deep, technical programming takes a while. That stuff isn’t worth learning unless you find it particularly interesting. Building an entire project usually takes some amount of non-trivial work, so I’m also going to say don’t learn programming to build one project. Spend your time validating the idea and using that evidence to convince yourself, a co-founder, or an investor that it’s worth putting some money behind.
However, you don’t need to know very much before you become programming-literate and start launching little webapps and working with APIs. And that can be very worthwhile for the non-technical founder for 2 reasons.
First, coding wildly increases the range of business and product idea which you can validate yourself. Without being able to put up a landing page and collect emails, the only tool you’ve got is customer conversations. Those are powerful, but they restrict you to qualitative feedback for problem-centric, sales-driven ideas.
In this case, I’m not talking about building the whole product — just that you’re able to cobble together enough of a prototype yourself to start collecting evidence about whether it’s worth pursuing in earnest or not.
Even if you already have a technical team at your company, being able to run the experiments independently has real value. It removes the dev team as a bottleneck and stops you from diverting them from their primary tasks.
The second reason you would want to learn, is that programming allows you to build your own tools. An abundance of quality APIs means you can skip the technically challenging bits of almost any tool you would wan to build for yourself. How about receiving an SMS whenever one of your posts hits the hackernews frontpage? That’s two hours of entry-level programming roping together twilio, hnsearch and google app engine. Without programming, it’s either impossible or a $1000 invoice from a freelancer.
The ability to build tools is consistently undervalued. Once you start thinking of the web as pliable, you start finding ways to save time, make money, and deliver value everywhere you look. Plus, it’s pretty fun.
So if you need to build one particular project, don’t bother learning. It’s not worth the time tradeoff. However, if you see startups as a career throughout which you’ll constantly have to test new ideas, or if you want to put a secret weapon in your toolbox, take the leap.
 I’d recommend starting with Google App Engine. It’s the fastest way I’ve found to get a new site live, has free tiers to start with, and completely eliminates ever touching server/system administration.