When a reporter asked Daniel Ellsberg how many Vietnamese people had died during the war he helped manage, he stared at her for a long time before saying, “A million? Two million?” He was one of the most powerful men in that war and he could not name the number. That is not a personal failing. It is what happens when you ask a primate brain to hold something larger than its evolutionary experience.

The human brain did not evolve to think about billions. It evolved on the African savanna, where your relevant social circle was maybe 150 people, your hunting range was maybe ten miles, and the biggest threat you faced lived in the grass. Your ancestors who worried about whether the tribe had enough berries that week survived better than those who spent their energy thinking about population dynamics a century out. Evolution does not care about your ability to reason about exponential growth. It cares about whether you ran from the lion fast enough.

This mismatch between our intuition and modern scale shows up everywhere, but it is most dangerous in three areas where the numbers are genuinely beyond human comprehension. Climate change is the obvious one. When scientists say the atmosphere has absorbed 90 percent of the excess heat from greenhouse gases, that sentence means something to you intellectually but nothing viscerally. You cannot feel nine hundred quadrillion joules of energy sitting in the ocean. You experience a hot day and think “this feels bad” rather than “this is one ten-billionth of the actual problem.” The gap between what your body tells you and what the data says is so enormous that it creates a kind of moral paralysis. Nobody panics about something they cannot feel.

Nuclear waste works the same way but in reverse. A half-life of 24,000 years for plutonium-239 is not a number your brain can hold. It is longer than all of recorded civilization compressed into a single figure. When you try to think about storing that stuff safely for 100,000 years, your mind hits a wall and gives up. The result is that we have built essentially nothing to deal with the problem, while generating more waste every year. We are leaving behind a physical legacy that outlasts every civilization in human history and we cannot even agree on where to put it because the numbers involved feel abstract enough to ignore.

Pandemics revealed this same pattern in 2020. The first confirmed US case was mid-January. By mid-March, cases had multiplied by roughly a factor of ten each week. That is exponential growth and humans are terrible at recognizing it because we think linearly. When something grows from one to two to four to eight, it feels slow. It does not feel like the thing that will fill every hospital bed in America by April. The people who understood what was happening were the ones who could actually see past the small numbers and imagine the curve continuing. Most of us cannot do that without a graph, and even then we tend to underestimate how steep it gets.

There is something useful to notice here about what does work when you need people to care about big numbers. The most effective climate communication I have seen does not try to make you understand the scale directly. It translates the abstract into the concrete. Instead of saying “three degrees of warming,” it shows you a map of your own city with sea levels three feet higher. Instead of talking about carbon parts per million, it tells you that the atmosphere now has more CO2 than at any point in 800,000 years, which is longer than Homo sapiens has existed as a species. Translation works because it respects the limits of your brain while still conveying the truth.

I think about this when I read about AI capabilities. People argue about whether models are “smart” or “conscious” or “dangerous,” but almost nobody stops to consider that these systems are manipulating numbers at scales their creators cannot fully trace. The weights in a large language model number in the trillions. No single human has ever looked at all of them. We built something whose internal state exceeds our capacity to understand it, and then we wonder why the outputs surprise us. This is not a new problem, but it is the first time the scale gap has appeared inside the tools we use to think.

The lesson is not that we should stop trying to reason about big problems. It is that we need better scaffolding. Maps help. Analogies help. Breaking massive numbers into smaller chunks that your brain can actually process helps. The people who make good decisions about climate, pandemics, or nuclear waste are not the ones with bigger brains. They are the ones who have built external systems for handling scale that their internal intuition cannot manage on its own.

Your brain is an amazing thing. It can recognize faces, navigate complex social situations, and write beautiful music. But it was never designed to hold a billion in its head at the same time as a million. Accepting that limit is not weakness. It is the first step toward building tools that compensate for it.