Wait. Remember that you get only one chance to make an initial splash and attract the attention. First impression lasts. There's no reason to waste the opportunity on something that is not ready or does not work 100% flawlessly.
If you really want to test the system on live people, do a round of closed alpha. However if you want to demo the stuff, wait until it's ready.
(an update)
Hmm .. I should probably add that the above is based not only on a common sense, but also on my personal experience. The project I was involved with grew from zero to a couple of million users in 1.5 years out of a single public announcement. If the project were not as polished as it was at that point, I really doubt the viral propagation would've as massive.
I agree with huhtenberg. The number of people who left quietly is much more than that of people who give you feedback when you don't have the minimal feature set.
If you really want to test the system on live people, do a round of closed alpha. However if you want to demo the stuff, wait until it's ready.
(an update)
Hmm .. I should probably add that the above is based not only on a common sense, but also on my personal experience. The project I was involved with grew from zero to a couple of million users in 1.5 years out of a single public announcement. If the project were not as polished as it was at that point, I really doubt the viral propagation would've as massive.