(Catching up on external posts on which I wanted to comment)
Uncertain Principles: Reader Request: Borrowing Energy
This gets answered in terms of the uncertainty principle in empty space. One thing Chad does not address here (to be fair, it was not explicitly part of the question) is that this is often asked and answered in the context of tunneling: You have a barrier of height V and an particle with KE < V, and yet the particle can be found to be on the other side of the barrier some fraction of the time. The answer is given that the particle borrows energy to get over the barrier, but it’s OK because the time is short. Unfortunately, that's the cheap way out.
The problem here is that "borrowing" energy is being used to explain a quantum phenomenon with a classical analogy. By borrowing energy the particle remains a particle and jumps the barrier in a nice classical way — the only quantum weirdness introduced here is the Heisenberg Uncertainty Principle. It's the kind of explanation given to people who have not been exposed to very much quantum physics, and it's semi-reassuring: the weirdness is limited to this whole uncertainty thing but all else is right with the world.
One of the physicists who posts to SFN (known as Severian) once remarked
Wave-particle duality is really just something we use to frighten children and undergraduates.
Funny, but I think this is actually backwards. It’s more like a Disney-fied story we tell to reassure undergraduates and non-physicists. You have the Disney happy ending that the world is still classical and we have a particle, which distracts from the original Brothers Grimm ending that no, these really are waves we are talking about, and yes, quantum mechanics is that strange.
As one of my professors explained it: “Quantum mechanics is weird not because things behave in a statistical sense, it’s weird because they don’t.” (See for example the Wigner semi-probability distribution.)