1
The Evolution of 'Impractical' Mathematics
MATH002 Lesson 5
00:00
In 1940, G.H. Hardy famously wrote that Number Theory was a "pure" science—so beautiful precisely because it was utterly useless for war or commerce. He could not have been more wrong. Today, the very integers he romanticized form the cryptographic armor of the digital age. This lecture explores how we transitioned from simple recursive puzzles to the RSA cryptosystem.

The Paradox of Continuity vs. Discreteness

In the world of continuous logic (Calculus), we rely on rules like the product rule:

$$\frac{d(fg)}{dx} = f\frac{dg}{dx} + g\frac{df}{dx}$$

Or recursive integration for functions such as:

$$\int \log^n |x| dx = x \log^n |x| - n \int \log^{n-1} |x| dx$$

While elegant, these continuous structures are predictable. Cybersecurity, however, requires one-way complexity. Discrete mathematics provides this through the logic of divisors and primes, where functions are easy to compute in one direction but virtually impossible to reverse without a "key."

Building Blocks: Mathematical Induction

Before we can secure a network, we must master Mathematical Induction to verify the algorithms that handle our data. Take the Fibonacci numbers, $f_n$. We can prove identities such as:

$$\sum_{k=1}^n (-1)^k f_k = (-1)^n f_{n-1} - 1$$

and verify growth rates using Binet-style relations:

$$f_n = \frac{f_{n-1} + \sqrt{5f_{n-1}^2 + 4(-1)^{n+1}}}{2}$$

This discrete logic, combined with Base Cases, ensures that algorithms like Insertion Sort (Alg 4.2.3) or the Tromino Tiling Algorithm (Alg 4.4.4) function correctly as they scale to trillions of operations.

From Patterns to Security: The RSA Shift

Modern security leverages Randomized Algorithms and the divide-and-conquer technique. By using the Fundamental Theorem of Arithmetic—the idea that every integer has a unique prime fingerprint—we create the RSA cryptosystem. Unlike the continuous curves of calculus, RSA operates on the "jagged" logic of prime factors.

🎯 Core Principle
Number theory provides "trapdoor" functions. While a divide-and-conquer search (Alg 4.2.1) can find a name in a list quickly, finding the prime factors of a 2048-bit integer without the key would take longer than the age of the universe.