Fermi Paradox

The lunchtime question

In the summer of 1950, walking to lunch at Los Alamos, Enrico Fermi fell into a conversation about flying saucers with Edward Teller, Herbert York, and Emil Konopinski. The talk drifted to the feasibility of faster-than-light travel and the odds of interstellar visitors. Then, midway through the meal, after a silence, Fermi suddenly asked the question that has carried his name ever since: "Where is everybody?"

The force of the question lies in a simple comparison of two numbers. The Milky Way contains a few hundred billion stars and has existed for roughly 13 billion years. The Sun is a latecomer, only 4.6 billion years old; countless Sun-like stars formed billions of years before ours and could have hosted technological life with an enormous head start. Set against that vast span of time and that enormous reservoir of stars, the time required for even a single civilization to spread across the entire Galaxy is laughably short. The expectation is that the Galaxy should be conspicuously inhabited. The observation is that it appears empty. That mismatch — high expectation, null observation — is the paradox.

Crucially, the paradox is not "we have not found aliens yet." It is the much stronger claim that, on the most natural assumptions, the Galaxy should already have been thoroughly settled, re-engineered, and made visible by now — and it has not been. The silence is not merely an absence of postcards; it is the absence of the civilization that should be standing on our doorstep.

The colonization timescale: why "they haven't arrived yet" fails

The quantitative heart of the paradox is the diffusion-of-colonization argument, formalised by Michael Hart in 1975 and sharpened by Frank Tipler in 1980. Imagine a civilization that sends self-replicating probes (von Neumann probes) outward. Each probe travels to a nearby star, harvests raw material, builds a few copies of itself, and dispatches them onward. The colonization front advances as a wave.

The effective speed of that wave is set by the cruise velocity v and the dwell time τ spent building copies at each stop, separated by the typical interstellar spacing d ≈ 5 light-years:

v_front ≈ d / (d/v + τ)

t_fill  ≈ D_galaxy / v_front     (D_galaxy ≈ 100,000 ly)

Plug in deliberately conservative numbers. Take a cruise speed of just v = 0.001c — one thousandth of light speed, about 300 km/s, well within the reach of chemical or nuclear propulsion concepts — and even allow a long dwell of τ = 1,000 years at each star. The crossing time per hop is d/v = 5,000 years, so v_front ≈ 5 ly / 6,000 yr ≈ 0.0008c. The Galaxy then fills in roughly:

t_fill ≈ 100,000 ly / 0.0008c ≈ 1.2 × 10⁸ yr ≈ 120 million years

Faster, more aggressive scenarios (v = 0.01–0.1c, short dwells) collapse that to 5–10 million years. Either way the answer is the same order of magnitude: a few million to a few hundred million years. Compare that with the time available — billions of years during which older stars could have spawned a settling species — and the colonization timescale is negligible. If even one civilization in the Galaxy's entire history chose to expand, the Galaxy should be full. The corollary, called the Hart-Tipler argument, is stark: the absence of extraterrestrials here and now is evidence that they do not exist anywhere in the Galaxy.

The Drake equation: organising the ignorance

The standard bookkeeping device for "how many civilizations should there be" is Frank Drake's 1961 equation, written for the SETI workshop at Green Bank. It estimates N, the number of civilizations in the Galaxy currently emitting detectable signals:

N = R* · f_p · n_e · f_l · f_i · f_c · L

where the factors are:

Term	Meaning	Best modern estimate	Confidence
R*	Star-formation rate (stars/yr)	1.5 – 3	Well measured
f_p	Fraction of stars with planets	~1	Measured (Kepler)
n_e	Habitable planets per system	0.1 – 1	Roughly constrained
f_l	Fraction where life arises	10⁻³ – 1	Unknown
f_i	Fraction reaching intelligence	10⁻⁹ – 1	Unknown
f_c	Fraction that signal	0.1 – 1	Unknown
L	Lifetime of signalling phase (yr)	10² – 10⁹	Dominant unknown

The astrophysical front of the equation has been transformed in the last fifteen years. Before Kepler, f_p was a guess; now we know planets are the rule, not the exception, and that roughly one in five Sun-like stars hosts an Earth-size planet in its habitable zone. But the biological and sociological terms — f_l, f_i, f_c and above all L — remain unconstrained by data and can each plausibly span many orders of magnitude. Because they multiply, the output N can be anything from far less than one (we are alone) to millions (the Galaxy is crowded). Drake's equation does not predict; it factorises our uncertainty, and the Fermi paradox is what happens when the optimistic end of that range collides with a null observation.

The Great Filter

If naïve estimates say the Galaxy should be full and it is not, then something must be suppressing N by a colossal factor. Economist Robin Hanson named that something the Great Filter in 1998. It is whatever step on the road from "dead but habitable planet" to "expansive, visible, galaxy-spanning civilization" is so improbable that essentially nothing makes it through. The chain has roughly nine links: the right star and planet, the building blocks of life, self-replicating molecules, simple (prokaryotic) life, complex (eukaryotic) cells, multicellularity, animals capable of tool use, technological civilization, and finally a long-lived, expanding one.

The filter could lie behind us. Abiogenesis — the jump from chemistry to a replicating, evolving system — happened only once on Earth as far as we know, and may be astronomically rare. The single origin of the complex eukaryotic cell, via the endosymbiosis of a mitochondrion roughly 2 billion years ago, took over a billion years after life began and likewise appears to be a one-off. If either of those is the filter, then microbial life might be common but complex life vanishingly rare, and we have already passed the hard part.

The filter could lie ahead. If the early steps turn out to be easy — if we find independent abiogenesis on Mars or in Europa's ocean — then the improbable step that keeps the Galaxy quiet must be somewhere we have not yet reached: perhaps technological self-destruction (nuclear war, engineered pathogens, runaway climate change, unaligned artificial intelligence), or some hard physical limit on interstellar expansion. This is why Hanson and others have argued, with grim irony, that the worst news for humanity would be to discover thriving biology elsewhere: it would relocate the filter to our future.

The relevant numbers

The paradox is fundamentally an argument about scale, so the figures are worth stating concretely.

Quantity	Value	Note
Stars in the Milky Way	1 – 4 × 10¹¹	~10⁹ are Sun-like (G-type) in the disk
Galaxy disk diameter	~100,000 ly (30 kpc)	Sun sits ~26,000 ly from centre
Age of the Galaxy	~13.6 Gyr	Thin disk ~8–10 Gyr old
Age of the Sun	4.6 Gyr	Billions of stars are older
Mean star spacing (solar neighbourhood)	~5 ly	Sets the hop distance d
Galaxy fill time (0.001–0.1c)	~10⁷–10⁸ yr	≪ time available
Earth-size HZ planets per Sun-like star	~0.2 (η⊕)	Kepler statistics
Time from abiogenesis to technology on Earth	~3.8 Gyr	Long, possibly bottlenecked
Detectable radio "leakage" window	~10²–10³ yr?	L term; we may be quietening already

The single most important comparison is the bottom-line ratio: the time to fill the Galaxy (~10⁷–10⁸ years) divided by the time available (~10⁹–10¹⁰ years) is at most a few percent and often one part in a thousand or less. Expansion is, in effect, instantaneous on galactic timescales. There is no way to make "they're on their way" the explanation.

The leading resolutions

Proposed answers fall into three broad families, distinguished by which term they collapse.

• Rarity (they don't exist). N is genuinely tiny because one of the biological terms — f_l, f_i — is near zero. The Rare Earth hypothesis (Ward & Brownlee, 2000) argues that complex life requires an improbable conjunction of conditions: a stabilising large moon, plate tectonics, a Jupiter-class shield, a quiet galactic neighbourhood, and more. In this picture microbes may be everywhere but we are effectively alone among the stars. The Great Filter lies behind us.
• Self-destruction (they don't last). N is large at any instant of emergence but L is short: civilizations reliably destroy themselves or stagnate within a few centuries of acquiring technology, so the steady-state number of contemporaneous, expanding civilizations is small and the colonization wave never gets going. The filter lies ahead.
• Non-detection / non-expansion (they're here or hidden). They exist and may even be widespread, but we cannot or do not see them. Variants include the Zoo hypothesis (deliberate non-interference, Ball 1973), the assumption that advanced cultures turn inward and do not expand (sustainability solutions), signatures we are not equipped to recognise, or the simple fact that our searches have barely begun.

Famous arguments and observational tests

• Von Neumann probes (Tipler 1980). Self-replicating machines make expansion absurdly cheap — one launch seeds the Galaxy. Their absence in our own solar system, where they should already be, is the strongest form of the paradox.
• Dyson spheres / swarms (Dyson 1960). An energy-hungry civilization might envelop its star to capture its full output, radiating waste heat in the mid-infrared. The Ĝ infrared survey (Wright et al., 2014–2015) searched ~100,000 galaxies in WISE data for galaxy-spanning waste heat and found no clear K-III "Kardashev Type III" candidate.
• SETI radio searches. From Frank Drake's Project Ozma (1960) through the Breakthrough Listen program (2015–), searches have grown enormously, yet a 2018 "cosmic haystack" analysis by Jason Wright and colleagues showed the total volume searched is comparable to one hot tub drawn from Earth's oceans. The famous 1977 "Wow!" signal at the Big Ear telescope was never repeated and remains unexplained but unconfirmed.
• 'Oumuamua (2017). The first confirmed interstellar object provoked speculation about artificial origin; the scientific consensus is that it is a natural body (a fragment or nitrogen-ice shard), but it underlined how poorly we monitor our own neighbourhood.
• Hart's silence (1975). Michael Hart's original paper, "An Explanation for the Absence of Extraterrestrials on Earth," argued from Earth's apparent virginity that no expanding civilization has ever existed in the Galaxy — the first rigorous statement that the silence is itself the evidence.

Common misconceptions

• "The Galaxy is just too big to cross." No — that is exactly the intuition the timescale calculation refutes. Even at 0.001c, the Galaxy fills in well under 1% of its age. Distance buys time, but there is overwhelmingly enough time.
• "The Drake equation gives N ≫ 1, so they must be out there." The Drake equation gives whatever you put in. Its later terms are unconstrained over many orders of magnitude; it cannot adjudicate the paradox, only frame it.
• "We've listened and heard nothing, so they're not there." Our searches have covered a vanishingly small fraction of stars, frequencies, sky positions, and time. A null SETI result so far is almost meaningless as evidence of absence.
• "UFOs / 'Oumuamua are the answer." No claimed terrestrial encounter or interstellar object has survived scientific scrutiny as artificial. Extraordinary claims still lack the extraordinary, repeatable evidence the paradox would actually require.
• "It's the same thing as the Drake equation." They are complementary but distinct. Drake estimates how many civilizations there might be now; Fermi asks why — given even a tiny number ever existed across cosmic history — none has made itself visible. You can believe N is small and still owe an explanation for the expansion argument.