My unhelpful answer is that I usually begin an unproductive procrastination loop wherein I overanalyze different alternative programming languages, frameworks, libraries, and possible ways to build software. I read a ton of books and tutorials and then never build the thing. Oh and I spend way too much time here and on other websites.
My helpful answer is that occasionally I break out of this procrastination loop and just start making something I think would be cool. It doesn't have to be anything new or exciting, just something that will keep my interest or even better scratch my own itch.
One of the biggest lies we tell ourselves is that you shouldn't reinvent the wheel. This couldn't be further from the truth. Just imagine if people actually followed this advice in the technology space. Almost all of the software that you use on a daily basis is a reinvented wheel.
