Skip to main content

Risk cost

The cost of the harm a system can do, read through named scenarios (a typical case and a specific worst case) rather than a single averaged expected value, because AI risk is usually tail shaped and an average washes out the rare severe failure that matters most. The hardest cost to quantify and the most dangerous to ignore, because it is where a system with tidy operating numbers hides a catastrophic tail. (see Topic 8.4)

Defined in 2 GAGE programs, which carry 2 distinct definitions of it. The wording above is taught in AI Governance: Applied Mastery.

How each discipline defines it

The same term does different work depending on who is using it. These are the definitions as each program teaches them, unedited.

AI Governance: Applied Mastery

The cost of the harm a system can do, read through named scenarios (a typical case and a specific worst case) rather than a single averaged expected value, because AI risk is usually tail shaped and an average washes out the rare severe failure that matters most. The hardest cost to quantify and the most dangerous to ignore, because it is where a system with tidy operating numbers hides a catastrophic tail. (see Topic 8.4)

Engineering Judgment and Professional Formation

The estimable cost of carrying a hazard today, including insurance premiums tied to a task's hazard class, expected injury-claim and lost-work-time cost based on incident history, and regulatory exposure. The relevant status quo cost category for safety-driven robot projects, discussed in Section 3G's Hard Case 1.

Where it is taught

The exact lessons this term appears in. The first 7 topics of every program are free with a free account.

Terms it appears with

Not an alphabetical neighbourhood: these are the terms taught in the same lessons, ranked by how often they appear together.