Calculus Indeterminate Forms Explained With Limits

I cover biotech and genetics. I write about CRISPR and protein folding and the algorithms that are trying to predict which drug molecule will fit which receptor like a key fits a lock. So what am I doing writing about calculus?

Here's the thing: the same math that makes antenna signal formulas break down at exactly their most important point is the same math that sits underneath computational drug design, neural network training, and the models researchers use to simulate how a misfolded protein cascades into disease. Indeterminate forms aren't a classroom abstraction. They're a structural feature of how we model the world — and if you don't know how to handle them, your model lies to you right at the moment of peak importance. 🧬

That's what the DIBEOS video How to Master Limits in Calculus is really about, even if it doesn't frame it that way.

Seven ways math breaks — and why that's interesting

The video lays out seven expressions that look like math but refuse to behave like it: 0/0, ∞/∞, 0×∞, ∞−∞, 0⁰, 1^∞, and ∞⁰.

The DIBEOS presenter describes them as a series of internal conflicts. Take 0/0: the numerator insists the answer should be zero (zero divided by anything nonzero is zero), while the denominator is doing something forbidden — dividing by zero collapses arithmetic entirely. Two rules, one expression, irreconcilable. Or take 1^∞, which feels safe — isn't 1 times itself forever just 1? But the presenter points out that 1.001 raised to the power of a thousand is nowhere near 1. A tiny deviation from exactly 1, iterated infinitely, accumulates into something completely different. The intuition fails.

The framing the video uses is conflict: each indeterminate form is a battle between competing mathematical forces, and the answer depends on which one wins. I find that framing genuinely useful. Most math education presents these as "undefined" and moves on. What DIBEOS is doing is slower and more honest — the answer isn't undefined, it's undecided, and the way you decide it is to zoom in and watch what actually happens as you approach the problem point.

That zooming-in process is called taking the limit.

The sinc function and the moment the formula fails

Okay here's the part that got me.

The presenter walks through the history of diffraction — Thomas Young and Fresnel in the early 1800s, studying how light bends around obstacles and creates patterns of bright and dark regions. The mathematics of those patterns eventually produced the sinc function: sin(x)/x. Elegant, useful, ubiquitous in signal processing.

Engineers later realized radio waves from antennas follow the same wave mathematics. For a satellite internet antenna, the sinc function describes the strength of the signal in different directions — and the most important point, the center of the beam aimed directly at the satellite, is where x = 0.

Plug x = 0 into sin(x)/x. You get 0/0.

The formula goes blank at the exact moment you need it most. The antenna is pointed at the satellite. The math says nothing.

What limits let you do is ask: what is sin(x)/x approaching as x gets close to zero, even if it never actually reaches it? And the answer — which requires comparing how fast sin(x) and x each approach zero — is 1. The signal at the center of the beam is at maximum amplitude. The math works. It just required a different question.

This same architecture shows up everywhere I cover. In protein folding simulations, models evaluate energy functions that approach indeterminate forms at the boundaries of conformational space. In machine learning, loss functions can produce 0/0 during training when a model's predictions become too confident — a divide-by-near-zero that destabilizes the whole system. The fix, always, is the same move the DIBEOS video describes: step back, study the behavior near the problem point, find the limit.

The tools for when direct substitution fails

The video covers a toolkit for resolving these forms. Two worth knowing:

Combining fractions. If you have 1/x minus 1/(x²+x) and you substitute x=0, you get ∞−∞ — two infinities whose difference is completely unknowable without more information. Instead, you combine them into a single fraction using their least common multiple. The two warring infinities merge into one expression that, when you substitute x=0, gives you 1. Clean.

L'Hôpital's rule. When a limit produces 0/0 or ∞/∞, you can differentiate the numerator and denominator separately (not as a quotient — separately) and take the limit of that new fraction instead. The geometric intuition, as the presenter explains it, is that you're comparing the slopes of the two functions near the problem point rather than their outputs. If the functions approach zero at similar rates, the ratio of their derivatives tells you who's actually winning the race.

Both techniques share a logic: you can't solve the problem at the singularity, so you transform it into a problem you can solve nearby.

A brief history of the three numbers that break everything

Here's what I keep coming back to, and it has that same texture as the genetic-code history I love covering — humans invented the notation before they fully worked out the consequences.

Zero showed up as a placeholder in Babylonia sometime in the first millennium BC, according to the MacTutor History of Mathematics at the University of St Andrews. (And "placeholder" is the key word — it was doing the work of distinguishing 34 from 304, not yet acting as a number in its own right.) The Hellenistic zero appeared in astronomical calculation, and Hindu mathematicians eventually promoted it to full numerical status. When Hindu-Arabic numerals reached Europe, zero came with them.

The infinity symbol — that sideways figure-eight — came much later, introduced by the English mathematician John Wallis in 1655. Why that shape? Wallis never explained his choice, and the honest answer is: nobody knows for certain. The omega hypothesis is the most commonly cited possibility.

What strikes me is the gap. Zero as placeholder, zero as number, zero in arithmetic, zero in calculus — each of those is a different conceptual upgrade, centuries apart. We named the thing before we understood it. That's not embarrassing; that's just how abstraction works. You see a pattern, you give it a symbol, and then you spend the next few hundred years figuring out what the symbol actually implies.

The 0⁰ controversy (slow down here — this one's genuinely weird)

Among all seven indeterminate forms, 0⁰ is the one mathematicians still argue about. The base wants the result to be zero (zero to any power is zero); the exponent wants it to be one (anything to the zero power is one). Classic standoff.

But here's where it gets interesting: in combinatorics and discrete mathematics, defining 0⁰ = 1 is genuinely useful and arguably necessary. The DIBEOS presenter walks through two cases. First, m^n counts the number of functions from a set of n elements to a set of m elements — so 0⁰ counts the number of functions from the empty set to the empty set, which is exactly one: the empty function. Second, the binomial theorem — a formula for expanding expressions like (x+y)^n — produces a term containing 0⁰ when x=0. If 0⁰ were indeterminate, the theorem would break for that case. If it were zero, you'd get a wrong answer. Defining it as 1 keeps the whole structure consistent.

And yet — the same presenter shows three limit problems that each produce 0⁰ under direct substitution but converge to three different values. The algebraic definition and the limit behavior aren't the same thing. Both are real. They just live in different mathematical contexts.

This is the kind of tension that doesn't resolve cleanly, and I appreciate that the video doesn't pretend it does. Math education often presents itself as a sequence of facts-to-memorize. The 0⁰ case is a reminder that some questions are still genuinely open, or at least context-dependent in ways that require careful handling.

The reason I ended up writing a calculus piece is the same reason I cover the algorithm that tries to predict which genetic variant will cause disease: both depend on models that can go meaningless at exactly the wrong moment, and both require the same basic discipline — don't accept the undefined answer, zoom in, ask what's really happening near the boundary.

The tools change. The question doesn't.

Mei Zhang covers biotechnology, genetics, and the future of medicine for Buzzrag.