How physicians think and use math to diagnose and prognose disease will profoundly change with the rise of artificial intelligence. We will examine four eras of clinical decision making to understand how physicians think and how they and their patients will use artificial intelligence.

The first era started 40 000 to 160 000 years ago when humans first developed modern cognitive abilities. Hippocratic school of medicine represents the second era and the introduction of modern prognosis and diagnosis into clinical practice.

Math formally entered into practice in the third era. Canada has been THE world leader in this era – as typified by Dr. David Sackett's pioneering work to develop clinical epidemiology and evidence-based medicine. In their daily work, physicians consciously and subconsciously use Bayesian inference that make would Thomas Bayes proud. A disease with a minuscule prior probability can be teased out against the backdrop of much more common conditions with similar presenting symptoms and signs.

Intelligent machines, the fourth era, will build on the tradition of clinical epidemiology to further evolve prognosis and diagnosis. There are three main contributions of these machines: i) enhance the capacity of patients in decision making; ii) increase the amount of information that can be used to make decisions; and, iii) reduce the cognitive biases that have persisted in humans for the past 40 000 years.