§ Library

Books and bets.

Books and essays I reread. Podcasts I rewind. Talks I keep returning to. And a small ledger of investment theses I am willing to hold in public. Updated when something earns the room, the list stays short on purpose.

Investments

Theses, positions, where the convictions sit.

  • FEB 01, 2026

    Iris

    open

    Building, in stealth

    Computer vision becoming ambient is the unbundling. Iris is one shot at it.

    Thesis

    When optical AI becomes a daily product, the winners will be the ones with a sharp single use-case, not the ones with a model card.

  • SEP 01, 2025

    Escape

    open

    Building, in stealth

    The gap between 'AI can technically do this' and 'AI is useful in someone's hand' is a design problem.

    Thesis

    Most consumer creative tools optimise for power users. The next billion-dollar product optimises for the user who never wanted to learn the toolchain.

  • JUN 01, 2014

    Anirar 360

    open

    Founder, ongoing

    Communication for Development is unfashionable, sticky, and trust-priced. Good things to bet on.

    Thesis

    Most film briefs in the development sector arrive without a Theory of Change. Whoever brings one to the room is the partner that gets renewed.

  • JAN 01, 2013

    10 Minute School (early-stage support)

    closed

    Production partner & first host

    Studio Imagine shot the first 10MS video and ran their first server. Watching from a distance now.

    Thesis

    Cheap, effective, vertical-format learning would eat horizontal-format institutions. It did.

  • JAN 01, 2012

    Studio Imagine

    open

    Founder, ongoing

    The longest-held position. Compounds because the network and the standard compound.

    Thesis

    AI is not going to replace creative studios. It's going to widen the gap between studios that can use it and ones that can't. I want to be on the right side of that gap.

Not financial advice. These are positions and theses, not recommendations.