If you are wondering which to implement, my suggestion would be to see if one really sticks out as cool to you. Don't worry about making something to get a lot of money/girls/ice cream (I know, that was first on your list of desires wasn't it...). If you have several ideas and they all excite you... Before I finish let me emphasize this point -- they all excite you. You are a lot less likely to finish or even make significant head way if you don't care about the idea. Now back to your regularly scheduled post. ... talk to your friends, see which one they get excited about.
Not always, but typically if you have a good idea, it should only take a few sentences (at most a paragraph) to get others excited about the idea. This isn't 100% perfect, but it is a nice rule of thumb.
If all else fails, just pick one. One thing all good programmers, from the Knuths to the Zawinskis of the world, have in common is they all program/read code constantly.
Last but not least, good luck. If you feel like it, show us the results or (if we are really lucky) the code.
Happy hacking, pwp
For each idea is the idea easy to make it real? Yes? Do it now No? Analyze more (give point on difficulty, probability of success, enjoyment doing it, etc...)
Once you only have _hard_ project, take the one with most point, start doing it. Only one at a time. Learn from your mistake for the next project.