Am I a shell?
Find out whether AI is amplifying your skills or turning you into a shell. Six questions. Honest answers. The math does the rest.
Take the assessment →Blog
- A Theoretical View on 'Something Big is Happening' February 2026 A critical look at the viral 'Something big is happening' article. Examining the claims about exponential AI acceleration with SWE-bench data, and introducing Shell Theory as a more moderate alternative.
- Context Engineering: Should You Bother? February 2026 Should you invest time in Rules, Commands, Skills, and Subagents? This article applies a framework of LLM limitations to current practices to identify which are worth building and which are workarounds with expiration dates.
- The Future of Context Engineering February 2026 Context engineering exists because LLMs have limitations. Agentic models obsoleted prompt engineering; some future innovation will obsolete context engineering. This article maps the current limitations and identifies which innovation we need.
- Shell Theory: Solving the AI Amplification Paradox January 2026 A simple model explaining how AI can raise baseline output, disproportionately rewards high-agency users, while at the same time risking long-term skill erosion.