No. For permanent genomic changes you have to select what you want. For precise edits, there's usually a counterselection (so it takes two hits). The efficiency of the process is still low. You have to be willing to throw away a lot of cells to get precisely the correct one. One way to think of it is a big component of what CRISPR does is to make it easier to find the good edits (although it does also make good edits more likely).