Real World Math
The mathematics behind finance, energy, and AI, in three parts for adults who want to use it, each opening with a twenty-problem diagnostic that says where to begin. Part I, calculus, runs from a battery's best duration and a bond's duration to the option Greeks and the chain rule inside a neural network. Part II, linear algebra and optimization, covers portfolio risk, linear programming, gradient methods, and the matrix arithmetic inside AI models. Part III, probability and statistics, runs from Bayes' rule and fat tails to regression and how models are fit and judged. Each module ends with a Memorize This box and 16 to 20 practice problems with worked solutions; the hours include about five minutes per problem.
## Thinking Straight: Fallacies, Biases, and the Numbers That Fool Us
About 2¼ hours · Last updated October 2026
The errors behind bad decisions, each taught through a case where it cost someone something: the fallacies, with the tobacco industry's 'Frank Statement' graded line by line; correlation and causation, from storks to the hormone-therapy reversal; base rates and Sally Clark's '1 in 73 million'; survivorship, Simpson's paradox, and regression to the mean; the halo effect and Theranos's board; confirmation bias and the fingerprint that three FBI examiners confirmed; anchors, frames, losses, and defaults; memory, hindsight, and the planning fallacy; overconfidence, and the court cases built on citations an AI invented; conformity, groupthink, and the Challenger launch; and the replication crisis. Where the field argues, over the hot hand, the backfire effect, nudges, the Dunning–Kruger effect, the replication crisis, whether people are irrational at all, and whether debiasing works, each side's case comes before the course's reading. Every module has a worked example, from a positive mammogram to a reference-class forecast for a tram. Opens with a Who's Who and a Field Guide to every error, and closes with a checklist for any claim.