The cost is embodied in the regulation, but the regulation is (in most cases) really just codifying the cost. You can't bring a drug to market until you've first done trials that it doesn't have any serious negative effects (or, at least, that you know what they are and can disclose them), and then until you've demonstrated that it actually works. This is expensive to do, because it involves doing controlled scientific experiments on groups of human subjects.
The fact that it's expensive means that it's not possible to explore all of them and so profit-motivated companies pick the ones that will give the most return on financial investment, which may or may not be the same ones that will save the most lives, or cause the greatest overall improvements in the standard of living.
The important thing to remember is that if it was so easy for him to get these documents, then that also means that there are about a million other people with the same clearance level as him who would find it equally easy. What's the betting that none of those are Chinese agents? Especially given how many Russian agents we've learned were working for the NSA and CIA during the cold war.
People focus on Snowden's disclosure as if it's possibly giving information to America's enemies (or, at least, not-so-friendly friends), but any of them that doesn't have a completely inept intelligence agency of their own will already have the information he's released. It was only secret from the people to whom these agencies should be accountable.
without pushing the seatback back (which I never like doing if there is someone behind me, I think airlines should remove that option)
Why? If the person in front of me in a flight pushes their seat back, then it moves the bottom forward very slightly, so I get about half a centimetre of knee room, and it moves the (small) screen of the in-flight entertainment system closer to my eyes. The seats are designed not to be made more uncomfortable when the person in front of you leans back...
The second point is why socialised health care and insurance companies have advantages over individuals in negotiating for care. An individual probably won't need to go to hospital in any given year, and very few of the ones that will need to can predict what treatment they will need in advance. In contrast, you can statistically work out roughly how many people in a country will need what kinds of treatment, with quite high accuracy. Negotiating to pay for them all together puts you in a much stronger bargaining position.
The big problem with this debate is that it conflates a whole range of choices in a single socialised medicine vs private medicine debate. In reality, there are a lot of points on the spectrum, depending on:
The question of what role the free market plays is complex. Obviously, you can't have people who have just suffered a heart attack shopping around for the best value ambulance to take them to the best value hospital. But you can have, for example, a central government buying medical services for all citizens (which typically counts as socialised medicine in these debates), but having different medical centres competing for the business, especially if they're allowed to take private patients as well so that they can stay in business when they don't have the majority of the government contract.
80+% of the people in the US HAD a good health insurance plan
Nope, 80% has a health insurance plan.
For society as a whole, we single payer countries tend to see better results. But per person, the healthcare in the US is the best. Assuming you have a good health insurance plan.
The last sentence here is the important one. And it means that, if you are either wealthy, or have a good job and no preexisting conditions (especially the kind that would stop you working for a bit) then you're better off in the USA. Or, to put it more cynically, US health insurance is a great deal, right up until the point where you need to make a claim.
Luckily, it's not too difficult to build a CPU which performs fully homomorphic operations. The math isn't even beyond high-school level (it's about on par with RSA). The primary issue right now is speed.
The primary issue with homomorphic encryption is speed, in the same way that the primary issue with running Quake 4 on EDSAC is speed. No one has come up with a general-purpose homomorphic encryption scheme that doesn't also come with a slowdown that is so many orders of magnitude that you would get faster results doing the work on your mobile phone than in Amazon's cloud with homomorphic encryption. There are some special-purpose schemes for simple database queries, but each primitive operation you need (at least) doubles the total dataset size (and insertion / deletion times).
grep me no patterns and I'll tell you no lines.