Erik Kling

The most consequential technology decisions are often made before their strategic consequences become visible. This talk examines how infrastructure choices, vendor dependencies, data placement, and governance assumptions determine which options remain available later. It introduces the Architecture of Optionality as a practical way to distinguish apparent flexibility from genuine freedom of action—and shows leaders how to use the Exit Test to recognize where reversibility is disappearing before dependency becomes control.

AI governance is often framed as a question of policy, ethics, or model performance. But once systems can recommend, decide, trigger actions, or coordinate other systems, the central question becomes architectural: what authority has been delegated, where does it reside, and can it actually be recalled? This talk examines the infrastructure of delegated authority and shows why control is not demonstrated by granting permission, but by the ability to interrupt, override, reverse, and recover. It gives leaders a practical way to design AI authority before operational convenience becomes institutional dependency.

Efficiency is visible every quarter. Resilience often appears as idle capacity, redundant capability, or capital that could have been deployed elsewhere—until dependency is tested. This talk examines why markets and institutions systematically undervalue options they have not yet needed. It shows how to distinguish waste from strategic resilience, how capital and infrastructure choices affect freedom of action, and how the Exit Test reveals whether an organization has preserved the ability to change course.