Witold Reichhartthe author
Essays

The writing.

The newsletter at witoldreichhart.substack.com has two sections. One is about AI and the human, the other about governed intelligence. Newest first, with the date of publication.

AI and the Human

The section is at witoldreichhart.substack.com/s/ai-and-the-human.

  1. 22 October 2026

    Shouting into the Lake

    Why a cell, a child, and a company all stop trying when nothing comes back.

  2. 11 August 2026

    An Intelligence Made Entirely of Maps

    What Korzybski knew in 1933 about keeping a map true to its territory.

  3. 31 July 2026

    The Quiet Stranding

    Stranded assets, applied to people.

  4. 28 July 2026

    Slavery in the AI Age

    Cotton, ledgers, licensed dealers: what made binding a person pay, and what AI does to the same arithmetic.

  5. 26 July 2026

    The Tower I Mistook for a Lighthouse

    A watchtower built in 1587 to warn of slave raids, read on holiday.

  6. 21 June 2026

    Intelligence Was Never in the Machine

    Plato got there first, with no instrument at all. A developmental biologist is now giving the oldest idea in philosophy physical force.

  7. 28 February 2026

    The AI Industry Is Building Neurons When Organizations Need Synapses

    Intelligence lives in connections. Biology, ecology, music, and chemistry agree; most AI strategy does not.

  8. 27 February 2026

    The Shell and the Pearl

    On the principle that runs from physics through enterprise AI to the space between two people.

  9. 8 February 2026

    The Science Fiction Writer Who Saw AI

    Everyone quotes the founders. A writer from Kraków saw it decades earlier.

  10. 27 January 2026

    Where Do Future Seniors Come From?

    The medium solves the knowledge problem. It does not solve this one.

  11. 14 January 2026

    The Thermodynamic Wall

    AI's real limit is electricity. What happens when AI starts building its own power supply.

  12. 13 January 2026

    To Those Who Have, More Will Be Given

    The Matthew effect of AI augmentation: why the speed of repositioning matters more than capability.

Governed Intelligence

The section is at witoldreichhart.substack.com/s/governed-intelligence. Three of the essays are also kept as pages on this site, in full text, as first published; a fourth page, Dynamics Blindness, explains the paper of that title.

  1. 6 September 2026

    The Right Size of Intelligence

    Ant colonies, brains, and failed multi-agent projects point the same way: intelligence is a property of organization, and smarter agents will not make smarter systems.

    Also the page on this site

  2. 21 July 2026

    Guided Emergence

    Coordination without a venue. When agents transact across firms, the rulebook, the operator, and the halt are written into the contracts.

  3. 16 July 2026

    The Oracle Problem Has Been Answered Three Times

    Civilisations learned to act on answers they could not inspect. Each time the fix went around the decision and left the oracle alone.

  4. 14 July 2026

    The Emergence Layer

    Why AI needs ecosystem design. The most consequential dynamics now arise between autonomous systems, above any one firm's controls.

  5. 25 June 2026

    Human-in-the-Loop Doesn't Survive Volume

    The control everyone names and production quietly stops running. Human judgment belongs where the consequence is.

  6. 23 June 2026

    Model Risk When the Model Acts

    The revised US model-risk guidance of April 2026 excludes generative and agentic AI from its scope. What replaces it is a check on each decision.

  7. 26 May 2026

    AI Joins the Family

    An organization absorbs AI into its existing habits, and the habits decide what AI becomes there.

  8. 16 May 2026

    Spec Is a View

    For AI that acts under supervision, the specification is a view. The source of institutional reliance is the governed claim graph beneath it.

    Also the page on this site

  9. 15 May 2026

    Admissibility

    For a defined class of regulated AI work the binding constraint is no longer the model. It is whether the institution may rely on what the model produced.

    Also the page on this site

  10. 30 March 2026

    Epistemic Immunity: Why Knowledge Graphs Need a Defense Architecture

    Knowledge graphs are living systems, and a living system needs an immune function.

  11. 26 January 2026

    Build the Medium

    How organizations become intelligent in the age of AI: by building the medium their knowledge lives in.

Earlier writing

Articles published on LinkedIn before the newsletter, in the order they appeared, left as written.

  1. 21 June 2026
  2. 27 February 2026
  3. 24 February 2026
  4. 22 February 2026
  5. 2 February 2026
  6. 30 January 2026
  7. 11 January 2026
  8. 11 January 2026
  9. 7 January 2026