Are We Born Too Early to Be Real?

If humanity lasts as long as the Earth stays livable, the odds that you would be born this early are about 1 in 100,000. Given those odds, I'd sooner bet we're living in a simulation.

That isn't where I started. I started with an idle thought: if we really have millions of years still coming, what is the actual chance of being born now, so close to the beginning? The answer was strange enough that I kept pulling on it.

Two numbers and one assumption

The Population Reference Bureau estimates that about 117 billion people had been born by 2022. Its model starts in 190,000 BCE with two people, assumes 80 births per 1,000 people a year through 1 CE, and a life expectancy of about 10 years for most of human history.

I decided to use the simplest possible formula that fits. Births in a year are the population times the birth rate, so the total is that product summed over every year there have been people:

$$B = \sum_t P_t \, b_t$$

where $P_t$ is the population in year $t$ and $b_t$ is the birth rate that year.

Prehistoric birth rates are high because life was short - a population that holds steady with a life expectancy of about 12 years needs around 80 births per 1,000 people a year. My setup differs from PRB's - starting in 300,000 BCE with a thousand people, a population curve fitted through a dozen historical estimates, birth rates tuned to the same assumptions PRB uses. It lands at about 125 billion by 2026. That's the arithmetic reproducing their number, not a second source confirming it, but it shows the figure survives a different starting point.

The future takes one big assumption: we stop dying of age. I'm treating that as a question of when, not if - a civilization that could stop aging and lasts millions of years is unlikely to keep dying of it. When is anyone's guess. I'll use 2100 - plausible at the current pace of progress, but still a guess, and the far rows barely move if it comes centuries later. Until then births stay where they are, about 132 million a year. After that they have to fall, because a population that stops dying and keeps the same birth rate fills the planet fast. If accidents and illness still take about one person in a thousand each year, lives run about a thousand years, and a population of 13 billion needs only about 13 million births a year - a tenth of today's. No population forecasts, no fertility trends. If births end up lower, every date below just moves later.

Now line up everyone who will ever be born, in birth order, and pick one at random. The chance they land among the first 125 billion is 125 billion divided by everyone:

$$p = \frac{n}{N} = \frac{n}{n + \beta_1 T_1 + \beta_2 T_2}$$

where $n$ is everyone born so far, about 125 billion, $\beta_1$ is about 132 million births a year for the $T_1 = 74$ years until 2100, and $\beta_2$ is about 13 million a year for the $T_2$ years from 2100 to the sunset.

That's the whole method. It's an old one, known as the doomsday argument - Brandon Carter came up with it and John Leslie first put it in print, and J. Richard Gott found it independently and published his version in Nature in 1993, counting years where this one counts people. I got there on my own and only learned afterwards that it had a name 😅

How early is early?

Say humanity sunsets in 2100, before aging is gone. Another ten billion people get born, and being among the first 125 billion is a 93% chance. Nothing strange about us.

Push the sunset to 3000 and it's still 85% - the slow births after 2100 add only another 12 billion.

At 10,000 CE - a future shorter than the stretch since farming began - it's close to a coin flip.

Mammal species typically last 1 to 2 million years. Give us another million and we'd be a fairly ordinary mammal, and being born this early drops to about 1 in 100.

The hard ceiling for staying on this planet is the Sun, which brightens by about 10% every billion years. After roughly a billion more it boils the oceans. A humanity that lasts that long makes us about 1 in 100,000.

Sunset year	Total ever born	Chance of being born this early
2100	~136 billion	93%
3000	~147 billion	85%
10,000 CE	~240 billion	52%
~1,000,000 CE	~13 trillion	1 in 106
~1,000,000,000 CE	~13 quadrillion	1 in 105,000

How the totals add up: about 125 billion people born before 2026, then 132 million births a year until aging is defeated in 2100, and 13 million a year after that.

And that's the gentle version. If we never stop dying of age and births stay at today's level, every far row gets ten times worse - a million years becomes 1 in 1,050, the Earth's lifetime 1 in a million.

So why are we here?

If the sunset is far away, being here is improbable, and improbable things want explaining. The top rows of the table need none. For the bottom ones, I can count five:

• The end is near. Robin Hanson called it the Great Filter - whatever stands between dead matter and life that spreads and lasts. We don't know whether ours is behind us or ahead, and this arithmetic votes for ahead. It doesn't need to be extinction either. A collapse to a few million people does the same thing to the count - a small humanity can last a million years and still leave us unremarkable.
• Even fewer humans. The table already has births falling tenfold, and the far rows don't go away. Defeating aging alone doesn't get us out - for a long future to make us ordinary, humans would have to almost stop being born. That's a breaking point in itself, where classic human civilization ends and evolves into something else (posthumans) - and a count of human births stops describing what comes next.
• We're in a simulation of this era. If this era gets simulated many times over, most people who ever live through 2026 are copies, not originals. Then finding yourself here is ordinary, however long the real future turns out to be.
• The sampling is wrong. The whole calculation treats you as a random draw from everyone who will ever live. That's an assumption, not an observation.
• Luck. Someone has to be early - the first 125 billion exist in every possible future. At 93% that's no surprise. At 1 in 100,000 it's a weak explanation, and the only one that asks us to change nothing.

Who would replay this era, and why?

Nick Bostrom's simulation argument from 2003 says one of three things is true: humanity almost certainly dies out before becoming posthuman, posthumans almost never run simulations of their ancestors, or we are almost certainly living in one.

His simulators are our own descendants, but nothing in the arithmetic needs them to be. It could be anyone who made it past their own filter - maybe working from whatever remnants of us they found, maybe from nothing but the right starting conditions. A good enough simulation doesn't need a fossil record.

Why would anyone replay this particular era? I'm speculating, but if this is the edge, the stretch where humans turn into whatever comes next, it's the part of our story most worth coming back to - for our descendants, or for anyone who once stood at the same edge. I can see three reasons to come back:

• Research. Rerun a lost civilization to see what it was, or rerun your own to recover what it forgot - knowledge, or whole ways of thinking that didn't survive the transition. The way the Animus in Assassin's Creed replays an ancestor's memories to find what they knew.
• Survival. Kurzgesagt's We Found a Loophole to Survive the End of the Universe borrows an idea from Freeman Dyson: a mind running cold, in an artificial brain, spends far less energy per thought, enough to stretch a fixed store of energy over thousands of times the current age of the universe. A civilization that moves itself into a digital world to last that long has time to fill and a world to build, and its own history is the obvious thing to build.
• Entertainment. A favourite scene, watched or felt over and over - not from the audience, but from inside one of the characters. We already build crude versions of this in historical games. A civilization with time and energy to spare would build better ones.

The more pivotal the era, the more replays, and the more likely this is one of them. In that version the Great Filter isn't waiting ahead. We're probably standing in it, and somebody on the far side keeps rewinding.

Why this might be wrong

If the assumptions hold, the table stands - each row is just arithmetic. What's disputed is the next step, using it to guess which row we're in. Philosophers have argued that step both ways for decades. Ken Olum argues it cancels out entirely, and István Aranyosi takes the same tools to the opposite of this post: we survive, and we're not simulated.

The price of being here

The table follows from two numbers and one assumption - 125 billion people so far, 132 million births a year, and an end to dying of age. Every way around it costs something - the future, the random draw, or the reality of the keyboard I'm typing this on. Luck is free, and explains nothing.

I lean equally toward two of them, because of the odds - the simulation, and a future with even fewer humans on purpose, where we turn into something else. It isn't set in stone - tomorrow I might find better reasons to believe in the reality, with the Great Filter chasing us 🙂