This morning, I was staring at a stock chart. A small toggle at the bottom—Linear vs Log—caught my eye.

I clicked between them and watched the same price history redraw itself. In one view, the movements were measured in dollars; in the other, they reflected proportional growth. A move from $10 to $20 looked identical to one from $100 to $200.

That simple toggle made me pause. Why do we even have logarithmic charts? What was the original purpose of logarithms—and what do they tell us about how humans create tools to extend our own capabilities?

A Shortcut Born from Human Limits

In the early 1600s, mathematicians and astronomers faced a major bottleneck. Every computation—whether charting planetary orbits or calculating trajectories—required endless multiplication and division. It was slow, error-prone, and exhausting.

Then came John Napier, a Scottish mathematician who recognized the limits of human calculation. Instead of trying to outwork those limits, he reimagined the problem. What if multiplication could be turned into addition?

That question led to the invention of the logarithm—a mathematical function that compresses complexity. Large multiplications could be replaced with simple additions, and divisions could be replaced with subtractions. Later, Henry Briggs refined Napier’s work into base-10 logarithms, and together they built the first log tables—the computational engines of their time.

Logarithms didn’t replace human intellect. They amplified it. Astronomers, navigators, and engineers suddenly had a way to think faster, calculate deeper, and explore farther. It was a moment when humanity acknowledged its limits and then transcended them through creativity.

Why Stock Charts Use Logarithmic Scales

That same spirit lives quietly in modern stock charts.

A linear scale treats equal distances as equal dollar changes—so a move from $10 to $20 looks very different from a move from $100 to $110. But a logarithmic scale treats equal distances as equal percentage changes. A 100% increase always looks the same, no matter the starting point.

And that’s how markets actually behave. Growth in business and in nature is rarely linear—it’s multiplicative. Logarithmic charts help us see what’s really happening beneath the surface: the rate of change, not just the size of it.

It’s another way of aligning how we see with how the world works.

Innovation Begins at the Edge of Human Capacity

As I reflected on Napier’s story, I realized something deeper: the catalyst for innovation almost always comes from an understanding of our own limits.

Napier understood that humans weren’t built for repetitive computation—so he designed a system that made repetition unnecessary. Every transformative innovation follows this same pattern: first, we encounter a boundary; then we find a new representation that turns that boundary into leverage.

Today, we face new limits—of cognition, data, and scale. And once again, we are responding with tools that amplify human potential. This time, that tool is Artificial Intelligence.

The Logarithmic DNA of Computing and AI

Logarithms aren’t just historical curiosities—they’re baked into the very fabric of modern computing.

  • Semiconductors are exponential by nature. The current–voltage relationship of a transistor follows an exponential law. Engineers use logarithms to linearize circuits, design amplifiers, and navigate vast ranges of current and voltage.
  • Algorithms rely on logs for efficiency. The Fast Fourier Transform (FFT) reduced signal-processing complexity from O(N^2) to O(N log N), making digital communication and imaging practical.
  • Hardware once used logs to simplify math. The CORDIC algorithm, used in early calculators, computed trigonometric and exponential functions using only shifts and additions.
  • AI itself is built on logs. Information theory defines entropy using logarithms H = -∑p log p. Deep learning models use cross-entropy loss functions and log-sum-exp tricks to stay numerically stable on GPUs and TPUs.
  • And the loop continues. Some AI researchers are now experimenting with logarithmic number systems (LNS) for chips—representing values directly in the log domain so that multiplication once again becomes addition, saving both power and silicon.

The story comes full circle. The same mathematical insight that freed 17th-century astronomers from arithmetic is now helping 21st-century engineers design AI chips.

The Philosophy of Leverage

What unites these moments—from Napier’s tables to modern neural networks—is a shared philosophy: admit the constraint, then change the representation.

We cannot compute infinite numbers, move infinite data, or think infinitely fast. But we can describe the world differently so that those limits become useful.
Logarithms turned products into sums. The FFT turned quadratic time into quasi-linear time. AI turns unstructured data into insight.

That’s the real art of progress—not the illusion of limitless power, but the wisdom of knowing where to apply leverage.

A Change of Scale

That tiny “log” toggle on a stock chart feels almost poetic now. It’s a reminder that progress often begins with a shift in perspective.

When Napier invented logarithms, he changed how we processed numbers.
When we built computers, we changed how we processed information.
Now, with AI, we are changing how we process understanding itself.

The goal, though, remains constant: to make the human experience richer—not by doing less, but by seeing more, with greater clarity, proportion, and meaning.

Innovation isn’t about escaping limits; it’s about finding the right scale where our limits become rails, not walls.
And every time we find that new scale—whether in math, in silicon, or in thought—we move forward.
One logarithm at a time.

Discover more from Startup Iceland

Subscribe to get the latest posts sent to your email.