It's probabilistic, so it could be any number of genes being changed. Assuming nothing environmental that would increase replication errors. There's nothing preventing multiple mutations, except that the resulting organism should still be viable.
Note that even if the number of mutations is small between two cell generations, they carry over to offspring. The neighbor's "naturally grown" seed may have hundreds or thousands of mutations compared to your seed, without any human intervention required.
CRISPR can theoretically change large sections. But we have just as much trouble as nature has trying to mutate while not killing the organism, or otherwise stunting its development or causing other issues. It's safe to assume most changes will be small in that context too.
Note that even if the number of mutations is small between two cell generations, they carry over to offspring. The neighbor's "naturally grown" seed may have hundreds or thousands of mutations compared to your seed, without any human intervention required.
CRISPR can theoretically change large sections. But we have just as much trouble as nature has trying to mutate while not killing the organism, or otherwise stunting its development or causing other issues. It's safe to assume most changes will be small in that context too.