My Three Years with AI

The source text, an article by Greg Twemlow, diagnoses the primary threat of widespread Generative AI use as Agency Decay, resulting from the tempting cognitive passivity of the "lukewarm cocoon." Twemlow rejects retreating from the technology and instead advocates that the Pragmatic Majority must adopt discipline, asserting that discernment is the 21st century’s crucial craft. To achieve this, he proposes the Context & Critique Rule™, a formal protocol designed to reintroduce necessary friction and ensure human sovereignty over AI-generated work. This method is based on the Cognitive Pareto, which suggests users must automate the 80% of routine tasks while personally authoring the critical 20% that shapes the ultimate outcome. The protocol mandates first defining explicit context before prompting and then rigorously critiquing the output, culminating in a recorded Decision Defence to guarantee the user owns the rationale for the final choice. This framework transforms AI into a cognitive expander, focusing the user's effort on strategy and logic rather than mere syntax, thereby ensuring Accountable AI. Read the article.About the Author - Greg Twemlow writes and teaches at the intersection of technology, education, and human judgment. He works with educators and businesses to make AI explainable and assessable in classrooms and boardrooms — to ensure AI users show their process and own their decisions. His cognition protocol, the Context & Critique Rule™, is built on a three-step process: Evidence → Cognition → Discernment — a bridge from what’s scattered to what’s chosen. Context & Critique → Accountable AI™. © 2025 Greg Twemlow. “Context & Critique → Accountable AI” and “Context & Critique Rule” are unregistered trademarks (™).