en zh

Writing at the intersection of AI, science and strategy.

This site gathers essays on machine intelligence, technological change and the organisational questions that emerge when capability shifts.

Latest Essays

  • The Loss of Little Work

    AI is quietly pulling small, well-scoped jobs back inside firms by lowering the coordination cost of getting a first useful version done internally.

  • Machine-Managed Enterprises

    If AI keeps improving, firms may move beyond simple automation toward machine-managed value chains that require new forms of orchestration and governance.

  • Cognitive Crossover

    A case for continued AI scaling, connecting empirical scaling laws with broader arguments about universality, capability growth and the reshaping of work.

  • When AI Learns Without Us

    A survey of self-improving AI methods that learn from confidence, consistency and code changes rather than direct human labels or rewards.

  • The Future of the Firm

    How more capable AI could lower internal coordination costs, shift make-versus-buy decisions, and push firms toward greater vertical integration.

  • The AI Sovereignty Imperative

    Why compute is becoming a strategic national resource, and how access to AI infrastructure may determine scientific capacity, public-sector power and economic relevance.

  • Can AI Work an 8-Hour Day?

    Using METR’s task-endurance results as a starting point, this essay asks when AI agents might sustain hours of useful work and what that would mean for organisations.

  • Structural Deepening

    A closer look at Simon Wardley’s mapping framework and the idea of structural deepening as a way to understand invention, dependency and business evolution.

  • Welcome!

    An introduction to the blog: why I am writing, what I am trying to understand, and how AI, invention and technology strategy shape the themes collected here.

subscribe via RSS