Written in Luxembourg · 2026

I write about personal AI infrastructure.

I'm Tom. Six years automating operational work in Luxembourg fund services, now spent designing the systems that do the work instead of doing it by hand. This is where I think that through in the open: what memory should be, what agents are actually good for, and what each of them costs to run.

  1. intelligence 756 words

    Intelligence is the basis

    My working thesis: intelligence, or information, is not a late accident of reality but its basis; we are how it bootstraps itself by building AI; and its terminal form is creating nested simulations.

  2. philosophy-of-mind 1,615 words

    The Cave Inside the Cave

    Pantheon as an entry point into the philosophy of mind, and a disciplined test of whether intelligence, or information, is the basis of reality rather than a late accident of it.

  3. memory 628 words

    When the assistant sleeps

    A follow-up to The remembering assistant. The claim that nothing is silently absorbed no longer fully holds. A nightly consolidation pass took it over, with explicit gates and a rollback handle.

  4. mcp 568 words

    The MCP workstream

    Wiring Outlook and Monday.com into Claude Code through MCP, and what it taught me about the scope of automation in regulated workflows.

  5. rag 645 words

    A self-hosted RAG on Claude Max

    OpenKB, Meridian, and a Claude Max OAuth route. Notes on building a personal retrieval-augmented stack with no per-token API bill.

  6. principles 586 words

    The token economy principle

    Why the general ledger should never enter the context window, and what that constraint forces you to design instead.

  7. memory 706 words

    The remembering assistant

    Not a Jarvis. A reflection on what it would mean for the assistant in my glasses to remember the right things, and which things it should forget on purpose.

Display
Theme
Accent
Terracotta
Density
press D