It is possible, but still difficult, to edit cells in-place. One example is the liver, a highly vascularized tissue (lots of blood vessels). This means that you can use a blood-based delivery method, e.g. blood infusion of CRISPR components + other compounds that allow these components to enter cells. Over time, more of the liver cells are edited and you can reach a reasonable proportion of the cells using this method using current techniques (40-90%). Note that for many diseases you would not have to reach 100% of the cells in order to have a positive effect on health.
There are a few labs working on better targeted delivery methods. In general the approach is to "wrap" CRISPR components in an envelope that is able to target specific cell-surface proteins for the cell type of interest.
Fascinating, thanks.
Basically an auto transplant with marrow
There's also more far-out ideas that might be viable involving permanently affecting the bodies biological machinery to improve it, like adding extra insulin producing capability to cells in diabetic patients.
Just to clarify, that is a real question, not rhetoric.