Where is everybody?

The story is that Enrico Fermi asked it over lunch at Los Alamos in 1950, in the middle of a conversation about flying saucers, and that the table went quiet. The question survives the anecdote because the arithmetic behind it is hard to escape. Our galaxy is about thirteen billion years old and about a hundred thousand light years across. A civilisation travelling at a thousandth of the speed of light, with no physics we do not already have, crosses it in a hundred million years — under one percent of its age. Settle, wait, send two more ships: even a slow, unenthusiastic expansion fills the galaxy in a geological eyeblink.

The surprise is not that we have found nothing. It is that one expansionist civilisation, anywhere, in thirteen billion years, would have been enough — and there is no sign of even one.

The arithmetic that makes it a paradox

Michael Hart put the argument in print in 1975 and called the silence Fact A: there are no intelligent beings from outer space on Earth now. His point was that Fact A is not a small observation. If interstellar travel is possible at all, and if any species anywhere has ever wanted to do it, the galaxy should already be full — so either nobody else has arisen, or something stops them, or they are here and hidden, which is a claim with the same evidential status as any other invisible presence.

That is what makes it a paradox rather than a disappointment. Every individual step in the chain looks plausible. Stars are common, planets are common, the chemistry is not exotic, the travel times are long but not prohibitive. The conjunction of all those plausible steps predicts a sky full of engineering, and the sky is quiet.

Drake's equation is a filing system, not an answer

Frank Drake wrote his equation in 1961 for the agenda of a small meeting, not as a prediction. It factors the number of detectable civilisations into seven terms: star formation, planets, habitable planets, life, intelligence, communication, and how long a civilisation stays detectable. The first three are now measured reasonably well, largely thanks to Kepler. The last four are not measured at all.

That is where the equation is routinely misused. Multiply seven point estimates and you get one number that looks like knowledge; the uncertainty has been thrown away in the act of picking the points. Anders Sandberg, Eric Drexler and Toby Ord redid the calculation in 2018 using distributions instead — ranges that honestly reflect how little the literature agrees about abiogenesis and about the emergence of intelligence, some of them spanning dozens of orders of magnitude. The result is not a different average. It is a different shape: a large fraction of the probability mass lands on there being no other civilisation in the observable universe at all.

If that is right, the paradox dissolves without any filter, any catastrophe or any hiding. The silence is what a universe with one civilisation in it sounds like, and the surprise was an artefact of multiplying confident-looking guesses. This is the most important thing to know about the Fermi paradox, and it is usually the last thing mentioned.

The Great Filter

Robin Hanson's 1998 framing is the one that stuck. Take the whole path from dead matter to a civilisation visible across the galaxy and break it into steps: the right kind of star, a habitable planet, self-replicating chemistry, simple cells, complex cells, multicellular life, tool-using minds, and finally an expansion that leaves a mark. The product of those probabilities must be small, because the sky is quiet. So at least one step is a filter — a place where almost everything stops.

The whole weight of the idea rests on one question: is the filter behind us or ahead of us? If the hard step was the origin of life, or the jump to complex cells — both of which took a very long time on Earth — then we are through it, and the emptiness is a fact about the past. If every step behind us turns out to be easy, then the improbability has to live somewhere else, and the only place left is in front of us.

Why a second biosphere would be bad news

Nick Bostrom's 2008 essay draws the conclusion most people find backwards: he hopes the search for extraterrestrial life finds nothing. The reasoning is straightforward once the filter is in place. Every discovery that makes an early step look easy — microbial fossils on Mars, biosignatures in an exoplanet atmosphere, an independent origin of life in Europa's ocean — moves probability out of the past and into the future, because the total still has to come out small enough to explain the silence.

It is worth being precise about what this argument is. It is a Bayesian shift, not a proof, and its size depends entirely on how confident you were beforehand. Someone who already believes the filter lies in the origin of life will be moved a long way by a second, independent origin in our own solar system. Someone who thinks the Sandberg–Drexler–Ord result is correct is not obliged to be moved much at all, because on that view the numbers never demanded a filter in the first place.

Where observer selection comes in

This is a site about the anthropic principle, and the Fermi problem is where observer selection does some of its most interesting work. We are trying to estimate how often life and intelligence arise, and our only data point is the planet we are standing on — which, necessarily, is one where every step succeeded. A biased sample of one cannot be used to estimate a base rate. However improbable the chain was, every observer in the universe finds themselves at the end of a successful copy of it.

Brandon Carter turned that into an argument in 1983. If the expected time for intelligence to evolve were much shorter than the lifetime of a suitable star, we should expect to have arisen early in our star's life; we did not, arriving after most of the Sun's main-sequence time. Carter read that near-coincidence as evidence for a small number of very hard steps in our own past. The argument is elegant, it is genuinely anthropic — it corrects for the fact that only successful planets have anybody on them to notice — and it has been argued about ever since. A 2025 reassessment by Daniel Mills and colleagues makes the case that the hard-steps framing bakes in an assumption about the Earth's habitability that the geological record does not support: if the windows for each step were opening and closing for planetary reasons, the timings stop being evidence about difficulty at all.

What would actually settle it

Unlike much of the rest of this site's subject matter, this one is empirical, and the searches are real. Breakthrough Listen has surveyed thousands of nearby stars for narrowband radio. The G-HAT survey looked at roughly a hundred thousand galaxies in the infrared for the waste heat a galaxy-spanning civilisation would have to dump, and found none where starlight had been substantially reprocessed. Transiting exoplanet spectroscopy is beginning to reach the molecules that a biosphere would put in an atmosphere. None of that is a null result strong enough to close the question, but each one carves a piece off the space of possibilities, which is more than any argument in this essay does.

What the toy on this site can tell you is nothing about any of it. The Universe Tuner stops at the point where a cosmos can support observers; it has no biology in it, no planets, no history after the first galaxies. The lottery counts universes that reach observers at all, not how many observers, or how far they get, or whether they ever build a radio. The Great Filter, if it exists, lives entirely in the part of the story this model does not simulate.