Is breeding crops really so different from editing their genes?

The two do the same kind of thing by different routes, and the biggest difference is how directly they reach the gene. Selective breeding, also called artificial selection, is how humans use animal and plant breeding to develop particular traits by choosing which individuals reproduce together. It has been practised since prehistory: wheat, rice and dogs have differed significantly from their wild ancestors for millennia, and maize needed especially large changes from teosinte. Genetic engineering is the modification of an organism's genes using technology, including the transfer of genes within and across species boundaries.

The old method is slow and works with what already exists. Breeders do multiple crosses and then pick the organisms with the desired traits. In the 18th century Robert Bakewell turned this into a scientific practice, and as farmers followed his lead, animals grew much bigger: the average weight of a bull sold for slaughter more than doubled between 1700 and 1786, from 370 to 840 pounds. Breeding can also be unintentional, and it can have unintended results.

Genetic engineering takes the gene directly from one organism and delivers it to another. That is much faster, it can use genes from any organism, and it avoids bringing in other unwanted genes. Between the two sits mutation breeding, which uses chemicals or radiation to produce many random mutations for breeders to select from.

Neither is risk-free, and both have critics. There is a scientific consensus that food from GMO crops poses no greater risk to human health than conventional food, while critics consider GM food safety a leading concern. And breeding has its own costs: dogs bred for extremely small sizes dislocate their kneecaps far more often.