Fermi Risk Agent explained

In 12 minutes, you will learn:

• How Enrico Fermi estimated the yield of the first atomic test from falling scraps of paper, and the habit behind his number
• Why risk matrices fail, and how three AI risks with the same score can sit a hundredfold apart in euros
• Why a chain of rough factors beats one guess at the total, and when it does not
• The anatomy of an AI risk estimate: records, measurements, assumptions and controls
• The rules and roles that stop anyone from steering the number
• How estimates add up to a register against a budget, and how weekly runs turn the number into a curve
• Why letting a chatbot place its own orders multiplies its risk by exactly five thirds
• Where the method stops, and what insurers ask before they cover AI risk