I dug around some more, and discovered to my delight it really is real. It goes by another name these days: magnetohydrodynamics. The study of motion of fluids under the influence of energy fields. If you remember the Red October's stealth propulsion system, that's the general idea.

It works like this. You take two electrodes, one big and one eensy, and you put a voltage difference across them. You keep them far enough apart not to get an electric arc, but as close together as you can. What you have just created is an asymmetrical capacitor, with the air as the dielectric.

It's a function of electrical charge that it tends to concentrate at locations with tighter radii of curvature. Also, on a curved surface, charge will always reside on the outside of the surface, not the inner. If you could charge an egg, for example, all those extra electrons could be found on the surface of the shell, and most of them near the pointy end.

You also need to know that when you have two objects with differing amounts of charge on them, there will exist an electrical field between the two objects. We imagine that as looking like a bunch of field lines traveling from the one object to the other. The closer the field lines are to each other, the stronger the electric field is at that point.

Now, looking at the asymmetrical capacitor from the second paragraph, a whole lot of field lines converge in the very small volume of the small electrode. Which means there's a boatload of energy in a very small space. So much, in fact, that electrons will leap off of the electrode and attach themselves to surrounding air molecules. If you're building a capacitor, that sucks, but for this purpose we want that to happen. Now, those air molecules are charged, and start to move along the field lines toward the other electrode.

This is the clever part. Not all the air molecules in the area get charged (ionized). They just hang out and do their own thing. Until an ionized molecule hits it. Then, it gets dragged along with the ion. Maybe it goes the whole way, maybe it doesn't. The point is that for as long as the neutral molecule is being moved by the ion, the ion (and by extension the capacitor) is exerting force upon it. And we all know what happens then, right? An opposite force pushes the capacitor the other way.

When the ion reaches the large electrode, it impacts the electrode and gives back any momentum it had received during the trip. The neutral molecules, on the other hand, keep going.

I should explain. Right now, you're probably imagining two solid disks, one over the other, one larger than the other. However, you could make the large disk a wire mesh. That would leave holes for the neutral molecules to pass through. Or, you could rotate the disks so they are edge-on to each other. While the ions turn sideways to impact the large disk at the end of their trip, the neutral molecules can just keep going straight.

Either way, the result is a non-ionized wind blowing through or past the large electrode. As you force the air down, the reaction force pushes the capacitor up. Do it hard enough, and it leaves the ground.

There are absolutely no moving parts involved. The only sound is the air moving. By the way, this is also how those ionic breeze air purifiers work.

Now, the downside.

Hobbyists make these all the time. It takes about half an hour to build a small model. It involves balsa wood and aluminum foil, and comes out about 20cm across, 8 cm tall, and triangular. Total weight in the dozens of grams range. They build a power source or rig one out of a computer monitor. The voltage used is about 30,000 volts, at roughly a milliamp. Stungun power.

These can just barely lift their own weight. The big challenge I read about was to break the 100 gram payload limit. That's a little less than a quarter of a pound. There has never been one built that can carry its own power supply. Which is, y'know, kinda necessary for a flying machine.

In order to be even marginally useful, it has to be autonomous. It also has to be able to lift more than its own weight by a significant margin. It has to be efficient enough that its power supply can be reasonably small while providing a decent flight time. It can't be made out of balsa wood and aluminum foil. That stuff's too fragile, and would never be able to support any kind of reasonable load.

The lifting capacity varies linearly with size, which is good. It means that if you can come up with Design A that lifts X ounces, you can put a number of them N together to get a total lift of N*X ounces. The problem lies more in the fact that lift is measured in ounces. A toy helicopter of similar size could power and lift itself easily.

The trick isn't to produce more ions. The trick is to get the ions you do produce to interact with as much neutral air as possible as it passes between electrodes. It's going to require an improvement in this area of at least an order of magnitude before anything approaching practicality blooms.

These are what we call "engineering problems." And you know I'm thinking about it.

UPDATE: The four most dangerous words in the English language: I've got an idea.