I mean, yes, that's how it would work if there were an infinite number of atoms in the piece. There's a finite amount, though, so eventually there will be a point when all the atoms have completely decayed.

All models are wrong, but some are useful.

Yeah, thats what I was using to get 14%.

2billion years is about 2.8 halflives, so I calculated (1/2)^2.8 ~ 0.14.

after a certain point, you're going to get to where you have to split an atom or two.

fairly sure that'd be far less exciting than normal.

Edit: i decided to try and figure out how long that would take.. and per usual the law of large numbers caused my eyes to glaze over.