George Carlin humorously argued that there's no such thing as the unnatural: if it can occur in nature, it's natural. The most artificial, synthesized, nanotechnological agent would still, in Carlin's eyes, be entirely natural. Why? Because nature allows for the possibility-- and thus the actuality-- of its existence. A thing whose existence has intelligence as one of its causes still falls inside the category of natural because intelligence is itself a product of natural forces.* By that logic, a unicorn, far from being a beast of the imagination, is entirely possible: genetic engineering might allow for such a creature (well, the creature itself at least, if not its magical properties) once we master the manipulation of horse and goat genomes.
The implications of this line of thinking are interesting to contemplate. Basically, all invention becomes a subset of-- and therefore a type of-- discovery. The car? A pattern that was discovered, and more efficient car-patterns are yet to be discovered. A novel? Why, a novel is merely the discovery of a certain pattern of evocative words. A musical composition? Notes in sequence, waiting to be found and instantiated. Mathematical principles? Math is the articulation of our discovery of the eternal apodictic realm.
I find the above train of thought charming, although I realize it has disturbing implications for my Buddhist sympathies. You see, the idea that there are patterns waiting to be incarnated smacks of Platonism, and I don't consider myself a Platonist-- I consider myself an empiricist. Empiricists aren't afraid to use inductive reasoning to divine abstract principles from concrete physical behaviors, but in the spirit of philosopher David Hume, they believe strongly that correlation does not imply causation, so what gets labeled as "causation" or "principle" is merely a regular concatenation of phenomena: event B regularly occurs right after event A.
I also have trouble with the notion of Platonic Forms: my own view is that, were nothing to exist, then the Forms wouldn't exist, either: nothing really means nothing. Of course, if we're to take the issue a bit more seriously, a more interesting question isn't whether the Forms would exist if nothing did: it's whether the Forms would exist if this physical universe didn't. My own gut feeling is that they would not: it's the brute physical fact of the universe that is logically prior to physical laws and other abstracta, not the other way around. Of course, I have no way to prove this, but I'm comfortable that the other side has no way to prove its case, either-- at least not in a universally satisfactory manner.
But is invention merely a subtype of discovery? Is it not a wholly different category of human endeavor? Doesn't it matter that mind plays a role in the process of invention-- that invention is a deliberate act of will whereas the term "discovery" implies accident, lack of volition? It's hard to say. So many of our scientific advances are an intertwined combination of accident and deliberate action. So many of our advances in mathematics result not from the relentless pursuit of logic but from intuitive flashes of insight. The word "discovery" itself suggests an uncovering, such that it matters little whether the uncovering is accidental or by design. So who knows? The natures of invention and discovery are fascinating to contemplate.
*I'm obviously taking a certain stance in philosophy of mind: the idea that mind is supervenient upon matter, an emergent phenomenon that is a function of material arrangement.