AI

How to approach AI

The author explores the transition from manual prompt engineering to a high-level "agentic" workflow where professional expertise is the primary driver of AI efficiency. A framework for delegation based on the Equation of Agentic Work is established.  This model weighs the value of automation by comparing Human Baseline Time ($T_h$) against the AI Process Time ($T_a$) and the Probability of Success ($P_s$), suggesting that as "Knowledge Work Effort" becomes an abundant resource, the true scarcity lies in the ability to provide "design intent" and recognize high-quality results. Ultimately, the "soft skills" of management—defining what "done" looks like and providing iterative feedback—have become the most critical technical skills for navigating the Jagged Frontier of modern AI.