Copernicus moved Earth out of the center of space. That was 1543. It took four more centuries for anyone to ask the obvious follow-up: does the same logic apply to time?

Gott asked it. His answer was the Copernican Principle, applied not to location but to moment. If you have no special reason to believe you are watching something at its beginning or its end, then you are probably watching it somewhere in the middle. The middle is bounded. Bounded means finite. Finite means the clock is already running.

This is not pessimism. It is a prior — a starting assumption before any other information enters the calculation. Gott called it the delta-t argument. You know one thing: how long something has already existed. From that single figure, you can construct a 95% confidence interval for how long it will continue. No biology required. No geopolitics. No threat modeling. Just duration, and the assumption that your moment of observation is not a miracle.

The machinery here is almost insultingly simple. If something has existed for a time T, Gott's formula says there is a 95% probability it will survive at least another T/39 years and at most 39T more years. The lower bound is the part that catches people. The upper bound is the part that comforts them. Neither is a guarantee. Both are claims about where you probably stand.

Applied to Homo sapiens — a species roughly 200,000 years old — the formula yields a median remaining lifespan of approximately 8,000 years. Not a catastrophe scenario. Not a tail risk. The statistically expected outcome, before any other fact about the world is considered.

That is what makes the number land like a weight.

What does it look like when the argument works?

In 1969, Gott visited the Berlin Wall. It had stood for eight years, since its construction in 1961. He did not run a formal calculation. He applied the logic informally, mentally. The result: a 95% confidence interval suggesting the wall would survive somewhere between a few more months and 312 more years.

The wall fell in 1989. Twenty years after his visit. Inside his predicted range.

One correct prediction proves nothing. Gott knew this. He said it himself. But the Berlin Wall case gave the argument something pure: a genuinely blind test, made before the outcome, with no cheating available. The interval was wide. But the world landed inside it.

Gott invoked this case for decades — not as proof, but as demonstration. The formula does not require special knowledge of walls, politics, or Cold War dynamics. It requires only that the observation be non-special. In 1969, he had no reason to think he was visiting the wall at an unusual moment. That assumption turned out to be correct.

The unsettling implication: if it works for walls, why not for civilizations?

How does the delta-t argument actually function?

It starts with self-locating belief — the question of what your own existence licenses you to infer. You are somewhere in the lifespan of whatever you are observing. You do not know where. In the absence of evidence that your position is unusual, the honest move is to treat it as random.

Random positioning within a lifespan generates a probability distribution. If something has existed for time T, and your observation falls randomly within its total duration, then with 95% confidence your observation falls between the 2.5th and 97.5th percentile of that duration. The algebra is elementary. The confidence interval falls out immediately.

The power is in what the argument does not need. It does not need a theory of extinction. It does not need to identify the specific threat — asteroid, war, ecological collapse, something not yet named. It bypasses all of that. The threat inventory is irrelevant. Only current age matters.

This is also the argument's most provocative feature. It claims to derive existential conclusions from almost no information. Critics — and there were many — found this either elegant or intolerable, depending on their philosophical priors.

Gott was careful about what he was claiming. He was not predicting doom on any particular timeline. He was establishing a prior. A starting point. A constraint on overconfidence. The formula does not say we will die in 8,000 years. It says that, before you add any information, the median expectation is 8,000 more years — and the 95% interval runs from roughly 5,100 to 7.8 million.

That interval is wide. Wide enough to feel almost useless for planning. Narrow enough to exclude civilizational permanence.

What counts as "humanity"?

This is not a rhetorical question. It is the central technical objection to Gott's framework, and it was pressed hardest by philosophers Nick Bostrom and Carlton Caves. The problem has a name: the reference class problem.

Gott's formula requires that you specify the population from which your observation is randomly drawn. The calculation looks clean when you say "all humans who ever lived." But you are not only a human. You are also a member of an industrialized civilization. A post-nuclear species. An entity alive after the invention of digital computation, artificial intelligence, and synthetic biology. Each of these framings defines a different reference class. Each reference class yields a different estimate. The formula gives you no way to choose between them.

This is not a minor technical complaint. It is a genuine structural gap. If you frame yourself as a member of all industrial civilizations that ever existed — a set currently consisting only of ours — the calculation becomes unstable. The sample size is one. Statistics require more than one.

Bostrom's Doomsday Argument, which shares ancestry with Gott's work, faces the same problem. So does every attempt to reason about self-location. The question of which population you belong to is not separable from the conclusion you reach. That circularity is real, and Gott never fully dissolved it.

He acknowledged the problem. He did not solve it. He argued that the most natural reference class — all humans who ever lived — was the appropriate one, and that the formula's usefulness survived the objection in practical terms.

Whether that is satisfying depends on what you want from a philosophical argument.

The 1993 Nature paper did not arrive from nowhere. Gott's career before it had established him as a serious structural cosmologist. His early work on cosmic strings — one-dimensional topological defects in space-time proposed as seeds of large-scale structure — was published in 1974 and earned him standing in theoretical physics. His 1991 work on the sponge topology of the universe, describing how matter is distributed on cosmic scales, was later supported by large-scale galaxy survey data.

This matters. Gott was not a philosopher gesturing at probability. He was a working physicist with a track record, applying a clean logical principle to a question that physicists had generally avoided. That combination made the 1993 paper harder to dismiss than it would have been from other quarters.

The paper catalyzed something. Bostrom, David Deutsch, Huw Price, and others built a serious philosophical literature around anthropic reasoning — the question of what existence licenses us to infer about our situation. Gott's specific numbers may be contested. The field he helped generate is not going away.

His parallel work on closed timelike curves — theoretical paths through space-time that loop back on themselves, potentially enabling time travel — shows the same disposition. Take a principle seriously. Follow it into uncomfortable territory. Name what you find.

His 2001 book, Time Travel in Einstein's Universe, brought that work to a general audience. His 2016 public prediction of the U.S. presidential electoral college outcome, made using Copernican-style reasoning about polling patterns, renewed attention to both his methods and their limits. It was a controversial application. It was also a demonstration that the posture — treat your moment as probably average, calculate from there — could be extended far beyond cosmology.

The delta-t argument is a prior. It is not a plan.

This distinction matters, and Gott was clear about it. Knowing that humanity probably has somewhere between 5,100 and 7.8 million years remaining — with 8,000 years as the median — does not tell you what to do on Monday morning. The interval is derived before any information about specific threats. It is a baseline, not a policy recommendation.

The formula cannot tell you whether the current century is exceptional. It cannot tell you whether the invention of artificial general intelligence, engineered pathogens, or nuclear weapons has permanently altered the shape of the distribution. It cannot distinguish between a civilization likely to run for its full statistical expectation and one that has already set in motion its own truncation.

This is precisely where critics like Bostrom diverge from Gott's framework. Bostrom's approach to existential risk is not Copernican — it is not built on the assumption that now is probably an average moment. It assumes that now may be an exceptional moment: the specific window in which a species either achieves or forfeits long-run survival. That assumption leads to very different conclusions and very different urgency.

The two frameworks are not easily reconciled. If Gott is right, then the most natural prior is that we will muddle through for a few more millennia at minimum. If Bostrom is right, then the Copernican assumption is precisely the wrong one to hold during a period of rapid capability growth in dangerous technologies.

What neither framework resolves is whether you can tell, from inside the moment, which scenario you inhabit. Gott's method requires that you cannot tell — that is the whole point. Bostrom's method assumes that certain structural features of the present make the answer visible, if you look carefully enough.

Both positions are internally coherent. The disagreement between them is not about math. It is about whether the present is ordinary.

Gott is not an eschatologist. He is not predicting catastrophe in a specific form, on a specific schedule, for a specific reason. He is doing something narrower and, in its way, more demanding: he is insisting that the rules of probability apply to us.

This is the move that generates resistance. Humans have always treated their own civilization as exceptional — too advanced, too interconnected, too aware of past collapses to repeat them. Every generation makes some version of this claim. The claim is almost always sincere. It is not obviously wrong. But it is not obviously right either.

Gott's argument is a corrective to that presumption. Not a refutation of it — the formula explicitly allows for civilizational durations in the millions of years. But a corrective. A demand that the burden of proof sit with exceptionalism, not with ordinariness.

The posture is scientific in the oldest sense: no special pleading. You are not excused from the distribution. The question is where inside it you are likely to stand.

That question has no answer that fits on a bumper sticker. It has no answer that comforts without distorting. What it has is this: a rigorous way to be uncertain. A framework for sitting with finitude that does not require you to know the specific mechanism of the end.

Gott looked at the oldest human fear and said: at least let us be honest about our uncertainty. That is not consolation. It is the beginning of clear sight.