For example, a study in that journal is entitled "Predicting blast-induced ground vibration using general regression neural network." The abstract is
Blasting is still an economical and viable method for rock excavation in mining and civil works projects. Ground vibration generated due to blasting is an undesirable phenomenon which is harmful for the nearby inhabitants and dwellings and should be prevented. In this study, an attempt has been made to predict the blast-induced ground vibration and frequency by incorporating rock properties, blast design and explosive parameters using the general regression neural network (GRNN) technique. To validate this methodology, the predictions obtained were compared with those obtained using the artificial neural network (ANN) model as well as by multivariate regression analysis (MVRA). Among all the methods, GRNN provides excellent predictions with a high degree of correlation.
Emphasis mine: they're testing if they can predict how the ground will shake after an explosion.
They're not going to spend five years dreaming up a model for it completely on a dry-erase board, set off a stick of dynamite, realize they were wrong, and throw out five years of work. They get preliminary results that are positive and encourage the project to go forward. They make up a model based on data, then refine it with subsequent bombs. If one were completely unable to use the method they propose to figure out how the ground shakes during explosions, they'd likely find that out before months had gone by.
I don't know what the breakdown is between clinical trials and basic research. I'd bet there's more research money spent on clinical trials than basic research (necessary, given certain realities about clinical trials) but that there are more scientists in basic research. So I'd bet that most research out there is actually not really affected a whole lot by the negative results issue: if you get a negative result, you get it before you commit yourself to it, and you move your reserarch in a different direction.
As far as why, the cocktail isn't super convenient. It's a bunch of pills taken throughout the day. Getting a toddler to take multiple pills a day every day is probably a very frustrating thing. I'd imagine the mother (who didn't have any prenatal care and didn't realize she was HIV positive until after birth) doesn't have the best health insurance, so the out of pocket expenses probably really quickly added up.
I mean, that's obviously all small potatoes compared to developing drug-resistant HIV and full blown AIDS and dying. Not excusing her actions, just saying it's understandable.
I'm not convinced it's a reasonable enough worry to say "lets not overhaul patents" though.
Thus, politicians gain very little and risk quite a bit by opposing the sanctions.
And yes, it is fucking stupid on multiple levels: it was probably always counter-productive, political leaders should show some fucking backbone and end it, citizens shouldn't be so apathetic about keeping an entire nation impoverished, and why is florida even allowed to vote?
But, dumb as all that is, "we still hate communism" is not a big reason why we still have sanctions.
And, again, any method of introducing crispr to a large amount of cells in the body would still be harder than just injecting someone with poison. There's no need to make it self-replicating like a virus.
In the double helix, when one strand of the DNA is broken, the string of DNA is held together by the other strand, it's an easy fix. Both strands broken at the same place means the DNA has come completely apart, has sustained some serious damage. The cell detects that pretty rapidly. The cleaner way is to find the sequence on the other chromosome and use that to repair the broken chromosome. A simpler but more dangerous way is to just grab two broken ends of DNA and stick them together. It's possible the cell will grab the wrong broken strand of DNA though and the result will be cancer.
Either way can be used to insert DNA where you want it.
Say a terrorist has a crispr combo that mutates several of your anti-cancer genes. He's got nothing: he'd need to get that into at least one of your cells in order to have any chance of giving you cancer. If he has a means to introduce it into one of your cells... he doesn't need crispr. He could just use a poison or some normal carcinogen.
Terrorists kill with pipe bombs and planes. Watching them try to do advanced biotech would in fact be quite hilarious.