What do an 18th-century Presbyterian minister, AI, conspiracy theorists, and good businesses all have in common?

Sounds like the start of a bad joke but it is one of the most important concepts of the 21st century, possibly of all time.

In the mid-1700s, Reverend Thomas Bayes gave the world a mathematical formula to describe how to update the probability of a hypothesis based on new evidence.

In other words: the ideal way to make decisions under uncertainty.

Bayesian statistics are the subject of some very passionate scientific discourse, even today. But there is little debate that our human brain - our intuitive, subconscious thinking - is very, very Bayesian.

We look at the sky and decide to take an umbrella. We start to worry after not hearing back from a family member. We cross the street based on the speed of oncoming traffic.

We have a prediction.

We perceive.

We update our predictions.

We perceive.

We update our predictions.

We perceive.

We . . . if you thought the next words were going to be “update our predictions”, congrats, you’ve discovered the link to artificial intelligence and LLMs. That’s (roughly) what they do, trained over enormous amounts of data.

That’s what Alan Turing did when we he cracked Enigma – that any given German sentence will contain E-I-N is a better prediction than X-Y-Z, even though they have the same odds if you’re checking randomly.

The degree to which you properly follow Bayes’ theorem – that is, the degree to which you properly pick the relevant prior beliefs and assess the base rate against your predictions, then adjust your thinking accordingly – describes how ideal your decision-making is.

Looking at dashboards and making a decision? Bayesian.

Fast moving startups: Bayesian?

P(A | B) = [P(B | A) * P(A)] / P(B)

The rest is commentary.