Bear in mind, though... "Anything nature can do" covers plenty of toxic, teratogenic, or carcinogenic results. The plants that evolved to kill us are ones we don't eat.

Blanket fear of "GMO crops" is obviously foolish, because the relevant question is "what genes does it have?" rather than "how did the genes get there?" But that doesn't rule out sensible concern about the higher rate of change on GMOs, and the ability to transfer potentially-dangerous genes wholesale.

The basic safety measure of natural selection and human cultivation is the rate of change. If a cultivar makes some people somewhat sick, it won't catch on or become universal, and it won't be intensified to something lethal. The real danger of GMOs is that they suspend that precaution - they can produce abrupt changes in an entire national crop. So the risks aren't different, but they might not be caught until they've done far more harm.

Sure, but on an evolutionary scale, humans and other dependent species have time to catch up to the changes. That's not true for changes wrought from GM.

Plant breeding has brought many adverse emergent properties. It involves selecting for a few alleles while having no way to prevent unintended alleles from being bred in. Introduction of invasive species have permanently degraded many ecosystems. The current method of mass-farming agriculture spreads purpose-bred plants around the world. The current method of agriculture is also high risk.

You can't speak of an "evolutionary scale" when there are literally intelligent designers at work, carefully cross-breeding each generation for optimal results.

It's a slower process than direct genetic manipulation, sure, but "evolutionary" implies thousands or even millions of years when the reality is closer to decades.

This is true enough, but even decades is a far different time frame from the average GMO. 30 years of directed breeding generally means a one-off cultivar being bred into something new and then popularized. The current GMO cycle involves a 1-5 years of test plantings, followed by national distribution.

So agreed, the evolutionary scale isn't applicable here, but it's still worth noting that GMOs represent a substantial acceleration from the directed-breeding scale.

Except "atomic gardening" is considered "organic", so triggering random mutations and seeing what works is much faster than "30 years of directed breeding".

https://mvd.iaea.org

It actually has already; invasive species choke out indigenous plants and permanently alter existing ecosystems.