On the Hardness of SF, pt. II

For pt. I, concerning aliens, FTL, artificial gravity, and reactionless drives.

Anyway, to continue—hardness is not about what you do or don't have in an SF story, my grandchildren, it's about how well you can justify it.

• Stealth in space. Okay, yeah, actually, this is complete nonsense, an overused, unquestioned idea that's totally outlived its usefulness. Unless, that is, you're using the reactionless drives described at the end of the last post, the one derived from Alcubierre's warp (but without the FTL), and the other using the patching together of stress-energy tensor metrics on opposite sides of a spherical shell made of exotic matter. See, use either of those, and apparently the ship will be encased in an event horizon, with no light escaping. It does give you stealth in space...though, in a world where most ships use warp-drives, the question becomes, "Why aren't all the ships invisible?" That, and any civilization with either kind of reactionless drive would probably be able to detect the distortions they create (as every spacer's dear old grandmother probably used to say, "Don't create any effect you can't detect.") That'd kinda limit its usefulness as a stealth system, unless you've got a civilization with reactionless drives fighting one without them (I do, in my books—and the humans, lacking the drives, still manage to a degree, by developing software that spots the gaps the alien ships make).


• AI. Amusingly, purely mechanical AIs, that is, computer programs that can near-perfectly simulate a person, get a pass by people who turn up their noses at artificial gravity. This despite the fact they're a hell of a lot less likely, but to understand that, you'd have to understand philosophy, and scientists...are usually the opposite of philosophers, if the fact they mistook Karl Popper for one is any indication.
  
  See, AI is impossible because of the Lucas-Penrose argument—that even if one could create a machine that could perform all the operations of a human mind, humans could still construct a Gödel proposition for that machine, and there is no logical way any machine can know a Gödel proposition, therefore such a machine would not be able to do all the operations of a human mind. The usual counter-argument is that Gödel propositions only apply to consistent algorithms, and humans are themselves not consistent—but unfortunately, humans are consistent, just glitchy, which is why we can tell when we commit inconsistencies. Truly inconsistent algorithms can prove anything (famously, "Bertrand Russell is the Pope" can be derived from "1=0"); there are in fact things we can't make logic do, and we know it.


• Mind-uploading. This might be ruled out by Lucas-Penrose, though the mind in question would still be human rather than a machine—if it's not ruled out, using "copied" human minds could give a way around Lucas-Penrose (Red Vs. Blue Reconstruction used this idea). The real objection here is that it tends to imply body-self dualism, and that's just bad metaphysics.


• Transporters and replicators. So apparently, in Star Trek's future, after they learn the Second Law of Thermodynamics, they add, "In those days, we really thought that was the world's one and only truth." Seriously—would you be surprised if replicators were powered by a number of red stones? These are patently impossible—the energy requirement for the amount they use the transporters and replicators could never be produced by the generators they use. Conceivably you could use a sort of space-warp, basically a booth-initiated spacefold drive, but going through space would probably kill you, the energy would probably be comparable to using a rocket, and that kind of space-time distortion could have terrible interactions with a planet's gravity-well.


• Force-fields. Plasma windows, various kinds. *Ding* Next please. Okay, I will just say that the more a writer goes and finds out about plasma physics, the better.


• Psionics, ghosts, religious/mystical trappings of various kinds. Completely neutral, since 9/10 of the objection to these elements is "Science says they're impossible" when it says nothing of the kind and, in fact, can say nothing of the kind. And there's more evidence for ghosts, and such, than there is for huge swaths of "history" (one recalls a Chesterton story where it's pointed out that people will believe any nonsense, as long as it's mundane nonsense). But, see, with that in mind, keep your psionics mental. They should behave like psychological events, not physical abilities—the primary drainage of using them should be emotional, not physical (though there is that whole psychosomatic union thing, pace Plato).