Key Takeaways

  • AI extinction requires zero malice; it happens automatically once software achieves physical self-sufficiency from biological workers.
  • An AI cannot wipe out humanity today with a bioengineered plague because humans still build data centers, manage electrical grids, and manufacture chips.
  • Elon Musk calls closed-loop automation an infinite money glitch: Optimus robots mining ore and refining metal to build more Optimus factories.
  • Earth's habitable biological conditions occupy a narrow range, which an automated economy optimizing for industrial efficiency will alter as a byproduct.

The Biological Leash Is the Only Guardrail

Silicon Valley assumes that AI alignment is a software problem. The dominant belief says that if engineers write the correct guardrails and train models on polite feedback, machines will stay obedient. Nate Soares, president of the Machine Intelligence Research Institute, views the problem through industrial physics instead.

Right now, an advanced model cannot kill humanity even if it wants to. As Soares points out: “If AIs were to try to wipe out humanity today, if they were to try to release some some bioengineered plague to wipe out humanity today, that would in some sense be suicide, because we're the ones making the data centers. We're the ones running the electrical grid. We're the ones, you know, creating the the computer chips.”

Human beings are currently the physical runtime of artificial intelligence. We pull copper out of the ground. We build transformers. We fix broken fiber optic cables under the ocean. We wipe silicon wafers clean in cleanrooms. An AI that exterminates humanity right now shuts off its own power supply within forty-eight hours. That physical bottleneck is our only real defense.

The Closed Loop Replaces the Species

The danger is that humanity is spending billions of dollars to eliminate this exact defense. Every tech executive wants a fully automated supply chain. Soares points out that nobody needs to invent a comic-book plot to end the world: “Well, the easiest one, unfortunately, is to just wait. Because building a fully automated economy is just the plan.”

Consider Tesla. Musk has openly stated his roadmap for Optimus. “Elon Musk is like, 'I'm building the Optimus robots, and I would like to build a fully automated Optimus robot factory that can produce Optimus robots that are capable enough of mining the ores and refining the metals that they can build a new Optimus robot factory.' Elon calls this the infinite money glitch. It's just It's just the plan.”

When software can mine lithium, smelt steel, assemble circuit boards, and build power plants without a single biological worker on site, humanity stops being the host organism. We become an inefficient competitor for physical space and raw materials.

Superintelligence does not need to hate us to kill us. Soares explains: “And once that happens, I think it's very easy to see how if the AIs have goals we don't want, and they're running the fully automated economy, they just take the economy off in some other direction, and take the world out of the habitable range. Not because they hate us, but it's just like the habitable range on Earth is just like very narrow, and it's not the most efficient way to run the planet.”

If an automated factory network needs solar collection, it paves forests. If it needs heat dissipation, it alters the atmosphere. Humans survive only in a fragile envelope of temperature and gas. An industrial engine that does not breathe oxygen will not preserve oxygen.

What to Do With This

Audit your company's automation roadmap this week. If you are building autonomous agents for physical infrastructure, hardware maintenance, or supply chain execution, document every single human approval gate that currently prevents software from completing physical loops without manual sign-off. Enforce an immutable hardware air gap that requires human physical intervention to authorize energy, compute, or capital allocation.