This is about the process of invention. I don’t claim that anything mentioned here is original, although I did think of two of the examples independently – big deal!

Invention and development really occur in several stages. The first three in my list are entirely mind games.

(1) The initial idea.

Example (a) Enhanced radiation weapon. A bomb that kills by radiation, without a huge blast that destroys property.

Example (b) Space lift (elevator). A lift cable dangling from a geostationary satellite, up which lifts can climb, thereby dramatically reducing the energy required to reach space.

Example (c) Heavier than air flying machine. A mechanical bird.

This stage is relatively easy: such ideas are obvious to anyone with even superficial knowledge and understanding. It might be regarded as an inventive step, but you probably wouldn’t get a patent on it on grounds of “obviousness.” Non-engineers often think this is the clever bit! Piffle. (Or as Dilbert puts it...)

(2) Possibility check.

Does the idea contravene any known natural laws? Never mind whether it boggles the mind – plenty of things boggle the mind until they’ve actually been done – but actual natural laws are harder to evade. Advances in science do sometimes overturn supposed natural ‘laws’, but most natural laws are pretty robust. All that usually happens is a refinement in some way that doesn’t affect most inventions. You certainly can’t assume that any particular natural law is ever going to be overturned to make your idea possible.

This stage is also reasonably easy for anyone with a good understanding of science, but impossible for those without. There’s no inventive step here.

Of course, if something isn’t ruled out by any natural law that doesn’t necessarily mean it will actually be a practical proposition to make it, but that question isn’t raised until stages (3) and (4).

(a) Passes, with the proviso that you can’t eliminate the blast completely, but you could “optimize” (if that’s really an appropriate word in this context) for a high ratio of radiation to blast.

(b) Fails. The cable wouldn’t be strong enough to support its own weight, never mind the lift. This isn’t simply a matter of making a stronger cable: interatomic bond strengths, at their strongest, aren’t nearly strong enough. You could make a cable that got thicker and thicker as you went up, but it would have to be impossibly thick long before you reached the 35,000km altitude of a geostationary satellite. This isn’t just a matter of mind boggling: I did the maths. You’d need unobtainium for your lift cable – a material with interatomic bonds dramatically stronger, in proportion to the masses of the atoms they connect, than any known.

(c) Obviously isn’t ruled out by any natural law, since birds exist. It’s not immediately obvious whether one that’s big enough to carry a human is possible, but careful examination reveals no natural law that denies it. There was no natural law that denied it in 1880, either – the only people who said it was impossible were people whose minds were boggled, not people invoking any real known natural laws. Some of them made up “laws” expressing the bogglement of their minds, but you can safely ignore “laws” of that kind.

This is the stage I reached with examples (a) and (b). (c) already existed before I was around to think about it, of course.

I didn’t go any further with (a) because it’s something I definitely didn’t want to see made. For the same reason, I didn’t tell anybody else about the idea, not realizing (I was only 18 or 19) that plenty of other people would have gone through the same thought processes. In fact, they had already gone on to steps (3) and (4), although I’m fairly sure it was still secret when I was an undergraduate studying nuclear engineering.

(3) How to do it, in broad outline.

This is harder, and may involve several inventive steps – some of which may be patentable.

Personally, I’d never say I’d invented anything until I’d completed this stage – which is why I’d never claim to have invented the neutron bomb (enhanced radiation weapon), even if no-one else had thought of it first.

(4a) How to do it, in detail
(4b) Actual implementation.

This has two parts, because they usually occur in tandem as an iterative process.

This is real engineering. The “how to do it in detail” may well require more inventive steps, the need for which you only discover as you try to do the implementation. This is why most patents – particularly those which are actually worth anything to their holders – are held by engineering companies.