Langlotz.AI
Writing

Published thinking

These are not separate posts. They are one argument about why AI stalls inside regulated institutions, grouped by the part of the problem they address. Newest first in each section. Where a piece builds on an earlier one it links to it, so either direction works.

Governance

Who answers for an AI system, and what the institution would have to be able to show.

Your Controls Are Perfect. They Are Applied to a Fiction. Read → Explainability Is a Liability Shield Read → Certified Trust Depreciates Read → How Do You Know Your AI Governance Lead Is Not Just a Rubber Stamp? Read → The Shadow AI on Your Balance Sheet Read → AI Has Control Planes. It Still Does Not Have a GAAP. Read →
Operating model

What replaces approval-by-committee once agents arrive faster than meetings can be scheduled.

Whoever Ships the Road Governs the Fleet Read → The AI Operating Model: Repeatable Trust Read → Design It Against the Constraints or Design It Twice Read → Your Next Chief of Staff Is Not Human. Build It First. Read → The Model Was the Easy Part Read →
Strategy

What AI costs an institution and where the advantage actually sits, including the costs on no dashboard.

Stop Thinking of AI as Software. It's Infrastructure. Read → Europe Is the Third Pole That Cannot Act as One Read → The AI Stall Is a Talent Death Spiral Read → Europe Doesn't Have a Model Problem. It Has an Engine Problem. Read → Most AI Strategies Fail at Second Order Read →
AI literacy

What a board has to understand before it can govern any of this.

Your AI Literacy Program Measures Attendance Read →