How do you know you're good at something a priori without spending the 10,000 hours to find out empirically?
You do what statisticians do ... take a small sample and estimate.
Spend 20 hours each doing 10 different things, and then pick 2-3 of them to spend 200 hours doing. Then pick the one. While it may take 10,000 hours to be truly awesome, you can probably get a sense of being "good" at something in a much smaller amount of time.
What you do should be at the intersection of what you love and what you're good at.