The thought process for me goes with the splicing genes in, let's say you have some disease which kills a crop. And someone puts in something to make that crop resist that disease. So the thought goes it can't have naturally occured or the disease resistent strain would have flourished naturally and spread all on it's own. So where did they get the new strain from and is it safe? I just don't know. And I'm going to be eating it. At school, I remember learning about food production techniques over the years in history class. No one teaches anyone this stuff in school though I'm sure. Maybe people just need educating. If the scientific community puts it's stamp of approval on it, I'm sure that will be good enough for me.
The something you put in the crop is usually from the same crop type, or a closely-related crop type. Every so often it comes from a different organism, but that's not as common.
Most disease-resistant strains aren't produced through molecular techniques but rather through selection experiments. Basically, you grow a crop, expose it to the pathogen you want to breed resistance to, and wait to see what lives. You take the surviving plants (there will always be some that have innate resistance due to genetic diversity; no disease on Earth is 100% lethal to its target organisms), breed them, and expose the new plants to the same disease. Take the survivors, breed them, re-select. After several generations (depends on the ploidy of the plant) you'll have a seed stock in which virtually all (but not quite all - natural diversity) specimens have the resistance you're looking for. If all you want to do is grow plants of that type with resistance, you're done after this step - just keep using the same seed stock. Re-selection may be necessary every few generations.
This is how we breed disease or herbicide resistant plants. Once we have a 'pure' stock for that genotype, you can then set about sequencing it to determine what gene conveys that resistance (I just described a process that can take months or years in one sentence; this is NOT easy). When the gene is identified you then have to figure out how that gene is regulated (switched on or off). Finally, you sequence the entire coding region you're interested in.
Once you have that sequence, we then have several ways to get it into another organism (or different strain of the same crop). Electrical shock tends to work sporadically, viral insertion is better. But first you need copies of the original gene - sequencing only tells you what the gene is made up of, it doesn't let you make copies. So you take the cells from the original disease-resistant strain, break them up to release the DNA, and then apply a technique called polymerase chain reaction (PCR) to select and amplify the particular sequence you want (probably a few times at several thousand dolllars a shot over several days) to amplify that sequence. Then you either insert it into a viral vector (again, weeks of work in one phrase) or try electrical shock on the plant germ cells to get the gene into the new plant. Once it's there, you then have to grow the plant up and re-expose it to the disease to see if it worked. If it didn't, back up until you figure out where you screwed up along the way. If it did, breed the new plants and conduct selection until you have a 'pure' stock again.
All molecular techniques do is take a gene sequence from one organism and put it into another - often of the same species, just a different strain, sometimes of a different species. It's tedious, detailed work that can be frustrating as hell but the beauty is that one you have that gene sequenced and know which PCR primers to use, it can then be done on a mass scale much more quickly than producing an entire pure strain from wild stock.
This same type of selection/isolation/extraction/amplification/insertion/selection procedure is the basis of pretty much all modern genetics. It's how we identify disease-resistance genes, antibiotic-resistance genes, chemical-resistance genes, genes that convey resistance to cold/heat, and many other things. It's used constantly in research as well, and a similar process was responsible for restoring the immune systems of SCIDs patients.
I mentioned that all of the genes we use to convey resistances occur naturally (which is true), so you're probably wondering why all the plants haven't evolved to have that resistance. The answer is selection pressure. Where there is high selection pressure (e.g. the place gets the same lethal disease every season), the local strain will probably have a high-proportion of disease-resistance. This can become permanent in the population, but it usually doesn't. The reason is that resistant strains often have drawbacks (increased resource requirements compared to non-resistant plants, for example) compared to non-resistant plants. The lower the selection pressure, the lower the proportion of a population that will be resistant to that selection pressure.
Bacterial antibiotic resistance is probably the publicly most well-known example of this phenomenon. There is currently a bit of a crisis in health care when it comes to a nasty bacterium called Staphylococcus aureus. "MRSA" may be an acronym you've heard that stands for methycilin-resistant S.A. There are now very high proportions of MRSA in health care settings around the world. The reason is because of over-use of methycilin. This antibiotic was for a very long while the best treatment against stubborn Staph aureus infections, and it was heavily used. Now, ordinarily antibiotics work by killing the vast majority of a pathogen, including some of the lesser-resistant individuals. There will always be a small number of completely resistant individual bacteria left, but they typically are incapable of causing further infection because they are either out-competed by natural flora (bacteria that live in and on us) and die, or the population of natural flora prevents them from growing to levels that cause disease. This is why broad-spectrum antibiotics can be ugly - if you kill off the natural flora, you make it easier for pathogens to come back. Anyway, over-use of methycilin without patients completing their course of medication and in isolated areas (health care settings) has resulted in the proportions of MRSA compared to regular Staph aureus in those areas skyrocketing. There are now some hospitals where virtually the entire Staph aureus population is MRSA. That's really scary, because there is only one other antibiotic that can treat MRSA (vancomycin), and we are now seeing the development of VRSA populations. These individuals have always existed, but when you apply selection pressure (antibiotic), you eliminate individuals susceptible to it and the surviving population reproduces to replace it, thereby increasing the proportion of the resistant population.
TL;DR: Resistance genes are all naturally-occurring. We can influence the frequency of the occurrence by applying selection pressure. This allows us to isolate a population that all carries resistance. If we want to spend the time and money, you can then isolate what gene is responsible and consider transferring it to another strain or species. Sometimes this works, sometimes it doesn't, and most of that depends on how closely related the species/strains are and how complicated the resistance mechanism is. When it comes to chemical/antibiotic/pesticide resistance, it's usually only a couple genes (sometimes one). When it comes to disease-resistance, it's usually several genes working in concert, which makes the whole process much more difficult.