Chapter 2 - Math You Can't Escape (But Can Tame)

The Crux

You can avoid some math in AI. You can't avoid all of it. The good news: you don't need PhD-level math. You need intuition for a few key concepts. This chapter builds that intuition without drowning you in proofs.

The Math You Actually Need

Here's the honest breakdown:

Must-Have:

• Linear algebra (vectors, matrices, dot products)
• Probability (distributions, expectations, Bayes' rule)
• Calculus (derivatives, chain rule, gradients)

Nice-to-Have:

• Information theory (entropy, KL divergence)
• Statistics (hypothesis testing, confidence intervals)
• Optimization theory (convexity, saddle points)

Overkill-for-Most:

• Real analysis
• Measure theory
• Functional analysis

You can be effective without the third category. Let's build intuition for the first.