Writing

• Oct 2026How far should an ambient agent go?50 hypothetical scenarios testing when an ambient agent should observe, suggest, act, or refuse, and why the two misses were safety controls the model acted on because acting looked protective.
• Sep 2026Is Trust Built With Less?When agents start taking real actions, the credible move is not more sensing. It is saying what you will not do, and sticking to it.
• Feb 2026Atoms Beneath the AlgorithmsHow century-old industrial companies became AI's most critical infrastructure.
• Feb 2026Designing for the Agentic EraHow AI will reshape consumer hardware, and what it means for designers.
• Feb 2026The Invisible EmpireThe battles we watch are not the battles that matter.