The Last Mile to the Stars

People keep getting worked up about UFOs, and especially about the idea that the government is hiding contact. I mean, I personally think aliens should exist. So did Enrico Fermi, the Italian‑American physicist who won the 1938 Nobel Prize and built the first nuclear reactor under the University of Chicago football stadium. Fermi was interested in the possibility of intelligent life beyond our solar system, and saw it as highly probable. He just couldn’t understand why they hadn’t shown up.

Most space nerds will know the story: At a lunch at Los Alamos in the summer of 1950 with Edward Teller and a few other physicists, Fermi performed a quick calculation on a napkin—the kind of thing only an astrophysicist would do casually over lunch—then asked the question that’s haunted the field ever since: Where is everybody?

Fermi’s logic was straightforward: the universe is incalculably old, and even our local galaxy is enormous. Given that much time and that much space, even one ambitious civilization should have left a mark—colonized the Milky Way several times over, in fact. But we see zero evidence of this. The silence is just plain weird.

Because the Fermi Paradox is so elegant and its logic so clear, people have been trying to solve it for decades.

1996, Robin Hanson—economist at George Mason University, former AI researcher at NASA Ames and Lockheed—published an essay called “The Great Filter: Are We Almost Past It?” that addresses the Fermi Paradox by placing us on a kind of evolutionary‑technological timeline. Hanson lays out nine thresholds: abiogenesis, prokaryotes, eukaryotes, sexual reproduction, multicellular life, tool‑using animals, our current human stage, then large‑scale colonization and visible interstellar expansion.

The fact that we seem to be the only folks in our neighborhood implies that at least one of those steps is almost impossible to cross. And since humans have already cleared several, the real question is whether the Great Filter is behind us or ahead of us. If it’s ahead, we’re in trouble and don’t know it yet.

It may be that the resource extraction required to bring about advanced general intelligence (AGI)—the tech we imagine might help us reach other intelligent species or become the colonizing agent of intelligence itself—ends up destabilizing the biosphere before it can be realized. The tool eats the grid. The grid eats the climate. And we brown out before we can cross the threshold.

There’s plenty of data pointing in that direction. The International Energy Agency reported that data‑center electricity demand grew 17% in 2025, while AI‑focused consumption surged 50%. They project total data‑center demand doubling to about 3% of global electricity by 2030. By that same year, U.S. data centers will use more electricity than all energy‑intensive manufacturing combined.

These facilities are unbelievably thirsty. Two‑thirds of U.S. data centers built since 2022 are in high water‑stress regions. Global AI water use is projected at 4.2 to 6.6 billion cubic meters annually by 2027. A single ChatGPT query uses roughly ten times the electricity of a Google search. You can see how the demand scales.

Maybe the filter is Skynet. Maybe it’s nuclear war. Maybe it’s climate‑driven scarcity and political breakdown. No matter the threat, stepping on the gas in the hope that AGI will save us—while accelerating processes that undermine planetary ecosystems—is a risky bet. And that’s assuming AGI is even technically possible beyond the parlor trick of large‑language‑model pattern matching.

AI adoption thus far appears to follow a ten‑year curve. ChatGPT launched in 2022; by 2025 it had hundreds of millions of weekly users. Infrastructure moves on a thirty‑to‑fifty‑year curve. Transmission lines take 10–15 years. Nuclear reactors take 15–20. Water‑rights reallocation takes decades. Meanwhile, an unregulated capitalist system guarantees consumption will continue regardless.

There are possible ways through: enhanced geothermal; small modular reactors; efficiency gains like DeepSeek V3; perhaps a cultural shift toward smaller, on‑device intelligence powered by quantum processing. But all of this remains speculative.

As one of my high school teachers was fond of saying, “almost only counts in horseshoes and thermonuclear weapons.” That’s likely also true of technological singularities, which only exist for real if they are crossed. And in this timeline, at least, that hasn’t happened yet.

Where is everybody? Probably waiting for the grid to come back up.