This problem is not specific to 3D printers. Just ask Sheffield Forgemasters, who famously made an "oil pipe" for Iraq back in 1990. Oddly, the pipe was enormously strong and specified to very fine tolerances. Turns out it wasn't for oil.

What about the nasty isotopes which didn't come from "that old thing"? Don't you worry about those? That's my point.

Chernobyl was a horrific design even by the standards of the time. Modern designs are immune to the kinds of accidents which happened at Three Mile Island, Chernobyl, and even Fukushima.

Coal burning also releases more radiation than nuclear power does. If the same safety rules were applied equally to both types of power, nuclear would come out way ahead.

Absolutely. And tinkering pretty much requires a keyboard, so the majority of Android or iOS devices are physically unsuitable for that kind of use. Carry a laptop when you need it and a tablet when you don't. I can see the attraction of a slim, light tablet device when you just need to check emails or buy tickets, and although the prices seem rather high for such simple devices it's obvious that people are happy to pay. But no matter how good these tablets get there will always be tasks which need a better (and bigger) human interface, and that's where PCs are unbeatable.

Sure, but that's the nature of huge projects. They require huge amounts of money from numerous sources, and each source wants to make damn sure they get something in return for their investment. Could a more efficient management structure be imagined? Maybe. Could it ever actually be implemented? No.

They say everything is fine because if they don't the investors will lose confidence and the project will collapse. The actual lack of fine-ness causes many headaches for the scientists and engineers who have to deal with unrealistic planning, but somehow they usually find a solution. Think of it this way: the managers say what they say because they have more faith in their staff than they do in the politicians funding the project.

But you need the money before you start. Nobody would invest billions entering such a competition on the off-chance they might get their money back.

There's a boundary?

If the risk was obvious, I would agree. However I'm not sure that one's vulnerability when cycling is really appreciated - just as people didn't used to feel vulnerable when driving without seatbelts. Even a minor fall onto hard pavement can easily break bones, and if the broken bone happens to be the skull then you can be in real trouble.

A little while ago I took a tumble when a startled animal ran into my bike. I landed on my helmet, which cracked, and was dazed enough to earn a day in hospital. Later, when a road safety group visited my workplace, I got a chance to find out roughly what kind of impact I had taken. They took the remains of my helmet and hit it with a hammer on an undamaged area until it showed damage similar to the original fall. It required quite a serious blow with a heavy hammer. They then delivered a similar blow to a force-measuring stand, which indicated that the force delivered was far more than that needed to break a skull.

In short, even on an empty country lane an unlucky fall can kill you. Until recently I didn't know that, and I suspect lots of other people don't know it either.

The grade of the helium is entirely irrelevant. The helium used to cool superconducting MRI magnets is recycled over and over: it doesn't wear out, and impurities are automatically removed during the liquefaction process. Wasting "old" helium is just as bad as wasting fresh.

Put yourself in the shoes of the spy. Your choices are

Plan A: be conspicuous. Run a high risk of being stopped, but have a story and hope you can bluff your way out.
Plan B: be discrete. Try to avoid being stopped, although if you are caught you are doomed.

Opting for Plan A means you have to infiltrate a plane-spotting group. Then you have to persuade the group to go to the base you're interested in. When you get there you can only take pictures consistent with being a plane spotter, or your cover story is blown. And when you are stopped (which is certain, because photography is forbidden), you have to hope that none of your friends mention your recent membership of the group, or your curious interest in this particular base. That takes a lot of time and work, the information you can gather is constrained, and the risk is substantial.

Plan B is far easier, doesn't suffer from the limitations of Plan A, and if you do a good job of hiding your camera the risk of being caught is low. And yes, I have read too many spy novels.

This has happened before. A group of British plane spotters were convicted of spying and sentenced to 3 years in jail. It doesn't seem to have occured to the Greek military that real spies will be a little more discrete in their methods.

BBC news article

This is the best answer so far by a mile. The Raspberry Pi has all the hackability of the 8-bit machines we grew up with, but runs a modern operating system with modern programming languages. And of course comes with plenty of games.

In addition to the usual accessories, I'd recommend getting a dedicated screen so that the RPi isn't competing with the telly for attention. Other useful things to have include a case (transparent of course), a powered USB hub (for WiFi and storing data to USB keys) and a spare SD card with the OS installed on it. If he's interested in electronics, look out for the recently released "Gertboard" interface - it's more electrically robust than connecting directly to the RPi. And books. Lots of books, on paper.

The trouble is that (following your analogy) the hospitals are inside the restaurants.

Premium cost = average accident rate * average accident cost + insurance company margin

It's the same formula whoever's driving, and the coefficients are worked out in the same way for robots as for humans.