Elon Musk isn’t exactly a steely-eyed missile man. He’s a dreamer and a pioneer. During his recent keynote on making humans a multiplanetary species — beginning with colonizing Mars — Musk answered a question from the audience about who would be able to participate. “It would be basically, ‘Are you prepared to die?’ Then, if that’s okay, you’re a candidate for going.”

Lest Musk be written off as merely a dreamer, though, it’s worth remembering that he’s mustered billions of dollars to pursue those dreams. SpaceX, Tesla, and SolarCity are just the latest commercial endeavors designed to make these dreams reality. Or to put it differently — I think more how Musk would see it — to uncover something about reality that seems unrealistic.

Are We Living In a Simulation?

One of the more fascinating Muskian musings lately concerns human consciousness and technology. Musk believes we’re probably living in a simulation designed by an advanced civilization. And he believes the odds against this are “one in billions.”

His argument runs something like this: In our lifetime, interactive simulations have advanced from the likes of Pong to photorealistic, immersive virtual reality. If you assume any rate of improvement in that advance, even a small one, then games will eventually become indistinguishable from reality. Since we’re on this trajectory, it follows that the odds that we’re in a “base reality” are “one in billions.”

He also says that we should hope this argument is sound, because otherwise — super-advanced virtual reality not having been created yet — we’re probably going to cease to exist sooner than later.

The idea of living in a Matrix-style version of reality is laughable. Maybe in part because it resembles the type of damn-the-torpedoes Mars colony mentality. On the surface, both seem downright crazy. But if you’re willing to give a shred of credibility to the notion of a multiplanetary species, though — and many are — you might also be willing to concede that living in a simulation isn’t absolutely implausible. And if it’s not absolutely implausible, by Musk’s argument, it’s absolutely probable.

Musk Makes Big Assumptions

I think Musk’s argument for simulated reality has major gaps. For example, to assume that an advanced civilization would de facto opt to pursue virtual reality simulation as we’ve begun to do isn’t necessary at all. Maybe they’d have a much bigger interest in heightening actual reality, or maybe technical progress isn’t a built-in feature of civilizational advance — who knows. It’s also a big leap to assume that just because games have been able to represent objects more photorealistically and immersively than before, that the same trajectory presupposes an ability to entirely subvert an individual’s intuitive self-awareness. And depending on how Musk sees consciousness existing within such a simulation, it’s arguable that his idea of subjectivity itself is ultra-theoretical (somehow emerging from sufficient data points) rather than integrated with and inextricable from bodily experience, which is the best we’re able to guarantee empirically.

Musk’s analysis is the latest in a long chain of similar questions dating back to the ancients. Alongside his other feats, though — producing futuristic cars, building huge rockets to colonize other planets — his theory is especially modern, in that he subjects all empirical value to the power of rational thought. A condition like “Are you prepared to die?” just to see what’s possible is as modern as they come.

Forge Ahead, But Don't Forget

The point I’m driving at here is nuanced. (And it’s easy to lose a point like that when talking about a guy like Elon Musk.) We need dreamers like Musk to press the boundaries of human understanding. There’s something dignified about stretching the creative potential of mankind to explore new frontiers. And about harnessing rational thought to get us there. But we also need to be careful to check assumptions about our nature and its key features, especially when they’re made in service to extreme ends.

We shouldn’t forget that certainty isn’t purely rationalistic or disembodied. That’s a historical circumstance against which one can make a strong counterargument. For example — to take a Muskian tack — if the apparatus of rationality is mostly unchanged over the last few thousand years, and if certainty is merely rational, then the necessary format of certainty for all humans should inevitably be rationalistic. But it’s not. So maybe there’s something incomplete with that approach.

Andrew M. Haines is the editor and founder of Ethika Politika, and co-founder and chief operating officer at Fiat Insight.