It seems like the best way to get an idea going to just to think of a problem. However, even this doesn't seem very easy. How do you do when you try to think of problems to solve? How do you poke your memory?
One thing I've found helpful is to keep a notebook (either online or physical) with every interesting idea, unsolved problem, or irritating deficiency that I notice. Once you get into the habit of thinking "how could I improve this?" or "why does this suck so much?", I've found that I've begun to notice an abundance of potential problems to tackle. There's a big difference between just thinking idly about something, and taking the steps to write it down.
By using the (according to Larry Wall) First Chief Virtue of a Great Programmer: Laziness.<p>I'm pretty lazy and my mind has the tendency to space out into thought, so no matter what I'm doing (programming, yard work etc) I always look for ways to make it easier on me. A lot of times there isn't much to improve and you just have to keep plodding along, but sometimes you get really neat insights and ideas.<p>For example, some time back I was working as a web developer doing applications for program managers at a large company. After a few months of doing that, it turned out a LOT of the applications they needed were very similar, so after I finished my intern work term and went back to school, I made an application that writes other applications for my 4th year design project.<p>The idea behind it is that non-programmers can use a GUI to generate a certain class of applications automatically. Turns out that was a very good idea, since while I was writing the prototype Coghead (www.coghead.com) came out with their public beta and their software aimed to do the same thing I was doing (albeit in a different way).<p>I could give more examples, but being lazy is generally a good source of problems. :)
You can try to solve your own problems, and see where they might be problems for other people, too. If you've run out of problems, you can try to be either more curious, or more discriminating.<p>You can hold a tool in the back of your head, and try applying it to everything, to see if it might fit. This requires discipline. It is often much better to hold a problem in your head, and see if new tools might apply, Feynman used this technique to great effect.<p>Or you can try to solve other people problems. This requires insight, but it also requires empathy. This is a hard thing to cultivate. It can't be turned on like a switch, and it's difficult to empathize with those far from your life experience. But there are many people who've lived lives similar to yours -- if your friends don't have any problems in their lives, they are unusually charmed.
It's a mindset. Start thinking of everything as "something you can fix". Every time something annoys you or you think of something you wish existed but doesn't, or you think of a novel way of doing something, write it down. After a couple weeks you'll realize that you have a lot more ideas than you thought you did, but you were just (falsely) assuming that you couldn't do anything about it.
When I solve my own problems, I end up creating the next "emacs" for the problem area. It is <i>awesome</i>. But people on the web don't want emacs' they want notepads.<p>So don't do that I guess.
Pretty much everything is suboptimal. Consequently, if you dig into anything deep enough, you will begin to notice many of its problems. And depending on how you "dig," people usually will just tell you the problems.
Well, you encounter problems on a daily basis. Next time you encounter something that was poorly organized, poorly created or just down right poor, then solve the problem that made it poor.<p>That is ultimately how I came to the vision of <a href="http://www.propertystampede.com" rel="nofollow">http://www.propertystampede.com</a> . Ultimately I saw a problem of lacking innovation in the real estate rentals world. So I am solving it.<p>Chris
It doesn't work you like that: you can't just think of a problem out of nowhere (nor a research project, nor an idea for a poem). It's a lot easier to pay attention for a while and notice when <i>you</i> have a problem.