There’s a way of reasoning about the evolution of intelligence that doesn’t rely on fossils at all — just on timing and probability. It leads to an uncomfortable conclusion: the very fact that we exist, and existed when we did, may be evidence that intelligence is extraordinarily rare. This is the “hard steps” argument, and it’s one of the more rigorous and unsettling ideas in the field.

Carter’s Insight

In 1983, the physicist Brandon Carter noticed a coincidence that bothered him. The time it took intelligent life to evolve on Earth — roughly four billion years — is comparable to the total time the Sun will remain a stable, life-supporting star, around ten billion years. From a probabilistic standpoint, that near-match is strange. If intelligence were easy, it should have appeared quickly, leaving most of the star’s lifetime unused. If it were nearly impossible, it shouldn’t have appeared at all.

The fact that it appeared late — but still appeared — suggests something specific. Carter argued it implies evolution had to pass through a small number of extremely improbable steps, each so unlikely that it typically takes longer than a planet’s habitable lifetime to occur. We made it only because we got lucky on each one, and only just in time.

What a “Hard Step” Is

A hard step is a transition so improbable that, on a typical planet, it simply never happens before the star dies or the world becomes uninhabitable. Candidates from Earth’s history include the origin of life itself, the emergence of the complex eukaryotic cell, the development of sexual reproduction, and the rise of technological intelligence.

The eukaryotic cell is the favorite suspect. As noted by researchers like Nick Lane, simple life dominated Earth for some two billion years before complex cells appeared — a delay that looks exactly like a planet waiting a very long time for an unlikely event. If complex cells are a genuine hard step, then countless planets may host thriving microbial life that never advances any further, the great pause never broken.

The Great Filter

The economist Robin Hanson reframed this in 1998 with a chilling phrase: the Great Filter. Somewhere on the long road from lifeless planet to detectable galactic civilization, there must be at least one barrier so difficult that almost nothing makes it through — otherwise the galaxy would be visibly full of civilizations, and it isn’t. The silence we observe demands a filter.

The unnerving question is where the filter lies. If it’s behind us — if the hard step was abiogenesis, or the complex cell, something we’ve already passed — then we may be among the rare survivors, and the future is open. But if the filter lies ahead — if technological civilizations reliably destroy themselves or hit some wall we haven’t reached yet — then our prospects are grim. The same observation, cosmic silence, supports both readings.

A Recent Test of the Idea

In 2021, a team led by Andrew Snyder-Beattie applied modern statistical methods to the timing of Earth’s evolutionary transitions. Their analysis supported Carter’s basic logic: the pattern of how long each major step took is more consistent with intelligence being rare — the product of several improbable jumps — than with it being a likely outcome of evolution. The hard-steps model, decades after Carter, still holds up against the data we have.

It’s worth stressing what kind of argument this is. It’s not a claim from biology about mechanisms. It’s a claim from probability and timing, which is part of why it’s so hard to dismiss — and also why it can’t be considered settled. The statistics rest, once again, on a single planet’s history.

The Counterarguments

Not everyone is convinced. One objection is that the “hard steps” might not be intrinsically improbable but merely dependent on slow environmental change — the steps took a long time because the planet had to change first (oxygen levels rising, for instance), not because the biological event itself was unlikely. If the pacing was set by geology rather than chance, the whole probabilistic argument weakens.

Another objection is selection bias of a subtle kind: we could only ever find ourselves on a planet where all the hard steps happened to be cleared, however unlikely that was, so our own existence can’t be straightforwardly used as evidence about the odds. These debates are live and technical, and no consensus has formed.

Why It Matters for the Search

If the hard-steps model is right, the galaxy could be teeming with simple life and nearly empty of minds — microbes on a billion worlds, intelligence on almost none. That would explain the silence without requiring any catastrophe, and it would mean the search for intelligent signals is likely to keep coming up empty even as the search for biosignatures eventually succeeds.

It also raises the stakes for what we are. If intelligence really is the product of several one-in-a-billion accidents, then a technological civilization isn’t just unusual — it’s something close to a cosmic miracle, and a fragile one. That possibility is exactly what the search is built to test. Find another civilization, and the hard-steps pessimism collapses. Keep finding silence, and it grows harder to argue with.

The Anthropic Trap

There’s a subtle reason the hard-steps argument is so slippery: our own existence is not neutral evidence. We can only ask “why did intelligence take so long” from a planet where intelligence actually arose — which means we’ve pre-selected for an unusual outcome. This is the anthropic principle, and it tangles every attempt to reason from Earth’s history to general odds.

Consider: if technological intelligence requires several wildly improbable steps, then the rare planets where it appears would, almost by necessity, be ones where the steps happened to clear just in time, late in the star’s life. So observing that intelligence arose late on Earth might not tell us intelligence is rare — it might simply be what any observer, anywhere, would find, because only late-success planets produce observers to notice. Brandon Carter understood this, which is why his argument is statistical rather than simple. The trap is that we can’t naively treat “it happened here” as evidence about how often it happens, because we couldn’t be having this conversation on a planet where it didn’t. Disentangling genuine information from this selection effect is one of the hardest problems in the field, and it’s why thoughtful researchers hedge their conclusions so carefully. A single example, when that example is necessarily ourselves, resists clean interpretation.

A Tour of the Candidate Filters

If a hard step exists, where is it? Earth’s history offers several suspects, each with a different implication. The origin of life itself is one candidate — but life appeared remarkably early on Earth, almost as soon as conditions allowed, which argues against abiogenesis being the bottleneck. The jump to complex eukaryotic cells is a stronger candidate, given the two-billion-year delay and the apparently singular endosymbiotic event behind it.

Other possibilities sit further along: the evolution of nervous systems and large brains, or the specific leap to cumulative technological culture, which happened only once despite intelligence arising many times. And then there’s the unsettling option that the filter lies ahead of us — that technological civilizations reliably collapse, destroy themselves, or stall before becoming detectable across the galaxy. Each candidate filter rewrites our prospects differently. A filter behind us means we’re rare survivors with an open future. A filter ahead means the silence is a warning we haven’t yet heeded. The evidence doesn’t currently single out which suspect is guilty, and that ambiguity is the whole anxiety of the idea.

What It Means for Our Future

The hard-steps debate isn’t only about the past — it carries a sharp implication for what lies ahead. If the improbable filters are all behind us, then technological civilizations, once established, might be reasonably durable, and our long-term prospects could be bright. But if a hard step still lies in our future — if the reason the galaxy is silent is that civilizations like ours reliably fail to survive their own technology — then the silence is less a curiosity than a warning written across the sky.

This is why some thinkers treat the absence of detected civilizations as relevant to questions of existential risk. The more confident we are that the early biological steps were hard (and so behind us), the more we should worry that the remaining filter is sociological and ahead of us — nuclear war, climate collapse, or some technology we haven’t yet invented. The argument is speculative, resting as always on a single example, and it could be entirely wrong. But it gives the abstract question of evolutionary bottlenecks an unexpectedly personal edge. Asking whether intelligence is an inevitable stage of evolution turns out to be, in part, asking how long intelligence tends to last once it appears — and whether we are early in our story, or near its end.

An Answer Only the Search Can Give

The deepest frustration of the hard-steps idea is that it can’t be resolved from Earth. Every version of the argument leans on our single example, and a single example is consistent with intelligence being common, rare, or unique. The probabilistic reasoning is elegant, but elegance isn’t evidence. What would actually move the question is data from beyond Earth: a second independent origin of life would suggest the first step is easy; the discovery of complex life elsewhere would suggest the eukaryotic gate is passable; continued silence from technosignature searches would gently support the pessimistic reading. The hard-steps model makes the search for life not just exciting but decisive — the only way to learn whether the gates that shaped us are barriers the universe rarely clears, or doors that open more often than our lonely vantage point can tell.

SETIworld follows the science of evolutionary bottlenecks and what they imply for life in the galaxy — join the portal to track the hard-steps debate.

References

• Carter, B., The anthropic principle and its implications for biological evolution, Phil. Trans. R. Soc. 1983
• Hanson, R., The Great Filter — Are We Almost Past It?, 1998
• Watson, A.J., Implications of an anthropic model of evolution, Astrobiology 2008
• Snyder-Beattie, Sandberg et al., The Timing of Evolutionary Transitions Suggests Intelligent Life is Rare, Astrobiology 2021
• Waltham, D., Lucky Planet, Basic Books 2014
• Simpson, The probability of habitable planets, 2017