Here is a fact that no one quite believes the first time they see it. Take the number 2. Double it. Double it again. Keep doubling, 30 times. What you get is not a "kind of big" number. It is a number larger than the population of India — 1{,}073{,}741{,}824, a bit over a billion.

2^{1}, \;2^{2}, \;2^{3}, \;\ldots, \;2^{30}
2, \;4, \;8, \;16, \;32, \;64, \;128, \;256, \;512, \;\ldots, \;1{,}073{,}741{,}824

Each step doubles the previous step. There is no cleverness. Just doubling. And yet after only 30 doublings the number has blown past any physical intuition you had for it. This is what "exponential growth" means — not fast, but runaway.

The tower you can actually see

Imagine stacking 2^n grains of rice in a pile, one on top of another. 2^{10} = 1024 grains is roughly a tablespoon — you can hold it in your hand. 2^{20} \approx 10^6 grains is a full sack, about a million grains. By 2^{30} you have a billion grains, enough to feed a family for several years. By 2^{40} the pile outweighs an elephant. By 2^{64} — legendary, from the chessboard-and-rice story — it would take several planets' worth of rice to fill.

The visual of this is the point. Below, watch the tower bar for 2^n grow as you slide n from 1 to 30. Notice how much of the first half of the slider looks flat — the numbers are so small compared to the final value that the early steps barely show up on a linear scale. Almost all the visible growth happens in the last third. That is the signature shape of exponential explosion: silence, silence, silence — then a wall.

Growing tower of two to the n from step one to step thirtyA single vertical bar in the centre of the figure. A slider below selects the step number n from one up to thirty. At n equals one the bar is barely visible. At n equals ten the bar has grown noticeably taller. At n equals twenty the bar fills most of the vertical space. At n equals thirty the bar has run off the top of the figure entirely. A readout above the bar shows both the step number and the actual value of two to the n, which at step thirty reads one comma zero seven three comma seven four one comma eight two four. n=1 n=15 n=30 ↔ drag to change n
The same bar, plotted on a linear scale, carries almost all its height in the last few doublings. That is the shape of $y = 2^n$: the first $10$ steps barely show up against the final value, because $2^{10}$ is only about $0.001$ of $2^{30}$. When a process doubles, the *last* doubling always contributes more than *everything before it combined*.

Why the last doubling outweighs everything before

Here is a consequence you should feel in your bones. The n-th step of a doubling sequence is larger than every previous step put together.

2^n \;>\; 2^{n-1} + 2^{n-2} + \cdots + 2^1 + 2^0

Because the right-hand side equals 2^n - 1. So the last doubling is worth more than the cumulative sum of all the smaller steps.

Why this sum identity holds: the geometric series 1 + 2 + 4 + \cdots + 2^{n-1} can be computed directly. Multiply by 2 and subtract: 2 \cdot S - S = 2^n - 1, so S = 2^n - 1. That is just 1 less than the next doubling, 2^n. The "last step is bigger than all the others" fact is baked into the arithmetic of doublings.

Concretely: after 29 doublings you have 2^{29} = 536{,}870{,}912. One more doubling and you have 1{,}073{,}741{,}824 — more than double 2^{29}? No, exactly double. But it exceeds the sum of all the earlier stepped values by 1. The final doubling is the single largest event in the history of the sequence.

What 2^{30} looks like in real terms

Some rough comparisons to calibrate intuition.

Each step adds one doubling, and each doubling adds one factor of 2 — but a factor of 2 at this scale is a gigantic jump in absolute terms. At the scale of 2^{30}, adding a single doubling is the difference between the population of India and two Indias. At the scale of 2^{60}, a single doubling swaps you from one galaxy to two.

The log-scale view

If the linear picture feels deceptive — because the early steps are invisible — the log-scale picture is honest. On a log axis, each step up the exponent is the same distance, regardless of how big the number has gotten. 2^{30} is twice as far from 0 as 2^{15}, not 2^{15} times farther.

That is the point of log scale: it stretches out the small numbers and compresses the huge ones, so that equal exponent-steps look equal. It is the natural coordinate system for anything that grows by multiplication.

The one-line takeaway

You do not beat exponential growth with patience. You either get in front of it — start early, let your money (or your practice) compound — or you get run over by it. "If I study 1\% more every day, I am twice as good in 70 days" is a statement about doublings. 1.01^{70} \approx 2. 1.01^{140} \approx 4. 1.01^{365} \approx 37.8. One year of tiny-daily compounding turns a unit of effort into nearly 38 units.

That is the kind of growth this visualisation is trying to show you. Slow. Slow. Slow. Then the wall.

Related: Exponents and Powers · Roots and Radicals · Number Systems · Operations and Properties