Alright, context here, I've worked in the Agricultural industry, and have a fair respect for science and nature in equal parts. These views are my own.
Let me start with what we call GM's precursor (the correct term you're looking for by the way is Genetic Engineering - GE), Hybridization before I talk about Genetic Engineering itself. We've been practicing Genetic Modification for millennia though Hybridization. Hybridization is selective breeding for specific traits through existing natural pathways (cross pollination, the like) and is what humans have been doing with all forms of life for millennia. We've selectively bred horses, cows, chicken, pigs, dogs, sheep, foodgrains, and flowers (we have an exceptional variety of flowers nowadays, far more than would have existed without human intervention) amongst a very long list.
These hybrids can be exceptional in their capabilities, Norman Borlaug saved the world from famine by crossing Dwarf Rice varieties in Mexico with Indian varieties to create the Indian agricultural revolution (and delaying the Malthusian crisis for the world as a whole). The good thing is that hybrids can still breed with local varieties and sustain life (though there are exceptions such as mules which are infertile). The reason local breeding is important is due to the fact that nature ABHORS stagnation (The Red Queen dilemma), and if you to try and keep one variety of anything everywhere, one catastrophe anywhere (a virus, a new predator) would kill that thing everywhere. It's an easily broken (fragile) way of living.
However, our hybridization efforts, successful as they are are only being looked at in a fraction of a fraction of the larger timeline that is mammalian driven life on earth, and what we sow today, we reap in the future. Just look at selectively bred dogs, many (Beagles, Basset hounds, Pugs, Boxers) have severe genetic defects due to excessive inbreeding for traits, to the extent that some breeds are no longer viable (there's no room left in their pedigree to breed new variety, meaning the ones we have are the last we'll have). Therefore, I suspect our existing, non-genetically engineered actions will already show their costs while we're still alive.
Hybridization, our best bet before picking up the tools ourselves with GE, itself is a bit dodgy in long term sustainability, especially when corners are cut to make short term gains. Often to meet urgent demand (fads), breeding happens without care, and this really really messes up the gene pool for later, and that's just a tragedy. It takes about 20 years to hybridize a successful new variety of plants, in many cases it takes half that for a new pest to come and make that variety obsolete. This is still acceptable in the grand scheme of things to everyone except modern businessmen who need yearly growth in sales.
GE (Genetic Engineering) is where we pick up the tools and artificially modify the DNA of an organism, and sell it as our own. I'll be outright, I am not in favor of GE outside of laboratories for the next twenty years, at the least. However, since we've already had GE plants outside of labs since the 1980s, I'm late for that train.
I'll quote Wikipedia here:
"Farmers have widely adopted GM technology. Between 1996 and 2015, the total surface area of land cultivated with GM crops increased by a factor of 100, from 17,000 km2 (4.2 million acres) to 1,797,000 km2 (444 million acres). 10% of the world's arable land was planted with GM crops in 2010. In the US, by 2014, 94% of the planted area of soybeans, 96% of cotton and 93% of corn were genetically modified varieties. Use of GM crops expanded rapidly in developing countries, with about 18 million farmers growing 54% of worldwide GM crops by 2013. A 2014 meta-analysis concluded that GM technology adoption had reduced chemical pesticide use by 37%, increased crop yields by 22%, and increased farmer profits by 68%. This reduction in pesticide use has been ecologically beneficial, but benefits may be reduced by overuse. Yield gains and pesticide reductions are larger for insect-resistant crops than for herbicide-tolerant crops. Yield and profit gains are higher in developing countries than in developed countries."
The article sounds wonderful, and I'm sure has been actively curated to be that way. The problem hidden between the lines? Those yield gains are today, the costs are hidden and in the future.
I consider GE a problem for four main reasons
1) Precautionary Principle
The precautionary principle states that if an action or policy has a suspected risk of causing harm to the public, or to the environment, in the absence of scientific consensus (that the action or policy is not harmful), the burden of proof that it is not harmful falls on those taking that action. The reasoning behind this is that if what's working is fine, don't try to fix it with something that can go wrong (and screw up everything). It's extremely hard to prove GE is safe under the precautionary principle at this point of time, without long term (10+ year) studies. Companies don't want to spend the time (would you invest in a product that may or may not be able to be sold after 10 years?)
A very very common line of argument here by GE advocates is that well, planes are harmful (they crash, people die), if we applied the Precautionary Principle to planes, we wouldn't have airplanes. To that my counter is that one 'new' plane cannot infect all other ones. Life can. Fun fact, there are many, well documented cases of GM crops breeding with local varieties and spreading their genes into the wild. Once it's out of the lab, it's in the wild FOREVER. Planes can be recalled, life cannot. If just one thing goes wrong with one variety, you could end up with a worst case scenario of plants' natural heritage being destroyed. That's not safe for world food security. It's like a game of russian roulette where every time you win, you gain a little bit of money. The one time you lose, all those gains go to waste. It's not a game worth playing.
2) Monoculture
In the short run, GE plants can outcompete local varieties in terms of cost (GM Cotton in India was insanely helpful for Indian farmers, who were able to reduce the cost of pesticide sprays be 80% or better in some cases). This reduced non-GM acreage by astounding margins. In doing so, it shifted a previously diversified bounty of varieties into a few GM ones, which had the GM trait inside. This led to local varieties (which I will remind had been successful for thousands of years) to be abandoned, and a strong shift to monoculture. Monoculture is unnatural, and one failure in one place can lead to catastrophe everywhere for the species (and the life depending on it, in this case, humans)
?3) Red Queen
Remember the Red Queen dillemna (run to stand in the same place)? It still holds. GM Cotton's first variety is already obsolete, it now offers very few advantages in protection over local varieties. Why? Because new pests came up. Bollworms used to be the biggest cause of damage to Cotton, but with those dying off to GM, other insects had a chance to grow, and now those are destroying the Cotton crops. You can solve one thing with GE, but then the rules will change. The problem is, that by then you've lost your old varieties, and are now a slave to playing an arms race with nature in order to stay alive. Previously, this happened on its own with hybridized varieties, but those were more resilient (and more of them, because monoculture wasn't a thing). So all we've done is make ourselves happy at good gains in the first few years, then be forced to out research nature herself (good luck) for the rest of our lives. Bad move.
4) Who benefits
GE is not something anyone outside a select handful of labs can do in the world. This is a good thing to the extent that it prevents random stupidity taking over as everyone rushes to GM their own crops to madness (again, it's never contained), but a bad thing as far as concentration of wealth goes. GE makes strong companies stronger, and that's a problem for food sustainability.
Final question: Do we need GE to feed the future? I'd say no, not yet, we still have major improvements to make in supply chain efficiency (preventing harvested food from spoiling on the way to your plate), and some more from hybridization and better farming methods. To go down GE would be forever committing us down the road of no return, of eventually modifying everything. As I've said above, I'm against this not because we can't, but because we're not wise enough to see the consequences of our actions till it's far too late (Hello industrialization driven global warming!).
I'm guessing this is probably more than you asked for, but by the same account it probably will answer your question for a long time (science can change, this answer could be obsolete in two decades)
