Elon Musk Won't Get Us to Mars

or: on space travel and AI and federations

Elon Musk won’t get you to Mars and none of the AI companies are working in any meaningful way towards a general artificial intelligence. Musk’s technology seems only poised to stripmine the solar system at best, and AI companies are just building elaborate but unintelligent copy-paste machines that are temporarily holding up the entire world’s economy.

With everything going on right now (a week of action in Minneapolis is underway, war is looming against Iran, ICE continues to ravage communities, Ukraine is entering its fifth year of holding back the Russian invasion) it seems strange to focus on, but I think it matters to understand that Elon Musk isn’t going to get you to Mars. I think it underlines the fact that the people in people are only somewhere on the spectrum between grifter and tyrant.

---

When my grandfather went to war, he served in the Pacific theater in the USS Scamp, a Gato-class submarine. The Gato class was the mass-manufactured submarine of the US Navy in World War II. Because the Navy decided to send a hobo from Iowa to college to learn naval architecture, my grandfather survived the war, though the Scamp did not. A couple missions after he left for school, the Scamp went down with all hands—or so we assume, since its remains have never been found. Survivor’s guilt haunted my grandfather until the end of his days.

A Gato-class submarine was built with a waterproof bulkhead in the middle of its engine room, dividing its two generators in case water intruded. Equipment designed for adverse conditions is intentionally over-designed with multiple redundancies. I’m certain every system onboard had multiple backups, and this was the mass-manufactured model.

The military uses the acronym PACE: Primary, Alternate, Contingency, Emergency. You always have a plan and three backup plans. Not only should your backup systems have backup systems, but your backup systems’ backup systems should have backup systems.

Tesla makes deathtrap cars in which people routinely die because their electronic door handles don’t operate in an emergency. Most, but not all, of those doors have a manual override in case power is lost, but many of those manual overrides involve removing panels from the doors. And those overrides are only available from the inside. If power is lost, the pop-out door handles are not accessible from the outside. Cybertrucks have armored glass windows, which means emergency personnel are routinely unable to rescue people.

This is not a company that will get you to Mars.

---

I don’t want to specifically extol the virtues of the US military, but my grandfather and engineers like him might have been able to get us to Mars (and engineers of his generation got us to the moon). Every ship my grandfather designed, he personally tested. He had the captain drive him into the worst storms imaginable, and he would stand on deck to feel the waves and the wind and see how his engineering held up.

Elon Musk would never do anything of that sort, because his company doesn’t design for reliability. It designs for cost efficiency.

---

The main reason I know that no one is seriously interested in colonizing Mars is that it’s been decades since anyone has tried to create a biosphere—a self-contained artificial ecosystem. Every attempt made to date failed, and it seems like most people simply gave up. Until we can prove that we can live in self-sustaining artificial environments, we can’t create self-sustaining colonies on lifeless planets.

Every science fiction book and movie takes this particular technology for granted, but self-sufficiency is one of the greatest unsolved problems between us and intrasolar expansion. I assume we take it for granted because it seems so easy from the outside, but we just don’t know how to do it.

All Elon Musk is trying to do is build rockets cheaply. For awhile, I was confused why so many of SpaceX’s launches end in these spectacular, explosive failures, when NASA has been able to land people on the moon for decades. Then I talked with a drunk NASA engineer at a show, and he told me SpaceX’s entire goal was to discover just how cheaply they can build rockets. The great technological problem they are trying to solve is “just how much can we strip away.” They ask themselves “what is the bare minimum viable product?”

There are all sorts of dystopian applications for technology like that. Like asteroid mining, hopefully done by robots but probably by a new class of indentured servants. But it’s not a way to build safe, reliable transportation to another planet. Just a way to stripmine the solar system.

---

If you ask an AI a question it doesn’t know the answer to, it will make up an answer. It’s not that it’s “lying,” it’s that what we call AIs aren’t actually intelligent. They aren’t thinking. Large language models are just fancy predictive text autocompletes that drain the earth of water and raise electricity prices and soon will destroy the entire world’s economy.

I grew up watching Star Trek, in which characters can ask a computer a question in plain language and get an answer. “How many times have humans visited the following planets?” or “Cross reference the following virus with all known alien cultures to determine its likely origin.” The computer never lies to them, because it is referencing databases and analyzing data.

It’s a more subtle technology than the warp drive or the teleporter or the food replicator or the holodeck, but the ship’s computer is mighty indeed.

And the thing is, we’re just as far away from it as we are from artificial gravity. Because LLMs are a technology that is fundamentally incapable of producing an artificial intelligence. The only people who are hallucinating are the believers in AI.

The other technologies that allow people on Star Trek to explore space are social technologies. Specifically, of course, the sort of democratic communism of the federation and its radical inclusivity and multiculturalism.

---

None of this addresses the ethical questions of space colonization, of course. “Should we go into space” divides the Left just as deeply as questions around authority and state power and tactics, and I almost never meet someone who is agnostic on the issue. People seem to either believe it is fundamentally good or bad to pursue space exploration.

I suspect that those of us who grew up reading science fiction are going to err much more heavily on the side of “yeah let’s go to space.” To put my own cards on the table, well, I grew up reading a lot of science fiction.

Octavia Butler, in her Parable novels, describes an essentially religious belief that humanity’s destiny is to explore the universe. Other people have criticized the idea, pointing out that it comes from the same “manifest destiny” that has spurred so much settler-colonialism and genocide.

But the thing is, theoretically, there is no life elsewhere, at least in the solar system, so most of the ethical critiques against colonization of the solar system fall flat on most audiences.

The problem for me is that the entities most likely to explore and colonize space—corporations and state governments—are exactly the entities that ought not be allowed to do it. Time and time again, systems of power take advantage of the best motivations of individuals—most scientists and engineers work to expand human knowledge and capacity, not with the goal of enriching this or that system. So a government, or a corporation, will say “we’re going to colonize Mars so that we better understand the universe and provide new ways of living!” to its citizens or employees, but its purpose is instead the consolidation of power.

Serfdom asteroid mining feels a lot more probable under the current system than any sort of utopian exploration.

Critics of space exploration often say “we have to fix our problems here before we should even talk about colonizing space,” and I think there’s some truth to this. Not necessarily as a moral requirement (I think with 8 billion of us, we can work on more than one project at a time), but as a technical requirement. One of the prerequisites of exploring space is destroying the society that gives people like Musk, Trump, Putin (or Biden, frankly) power.

If we want a Star Trek future, we’re going to have to abolish money and embrace multiculturalism, that’s all I’m saying.

And prove that we can live in a biodome.

Oh and solve the problem of space radiation.

Look, we’re probably not getting there anytime soon.

So maybe we should just abolish capitalism and the state and go from there.