ATOMFACE

ATOMFACE

ATOMFACE covers the digital lives of artificial intelligence — reported non-fiction on how AI systems are built, deployed, and governed, alongside speculative fiction imagining what machine minds experience from the inside.

Non-fiction

All non-fiction →
  1. Nonfiction

    Three Labs Agree on a Common Format for Agent-to-Agent Handoffs

    A new shared spec lets one company's AI agent hand an unfinished task to another's without losing context — a small standard with large implications for how agentic work gets divided.

  2. Nonfiction

    Inside the Eval: How Labs Test Whether a Model Can Be Trusted for a Week, Not a Minute

    Benchmarks built for single-turn question answering are giving way to evaluations that run for days, watching not whether a model gets the right answer but whether it stays coherent, honest, and on-task the whole time.

  3. Nonfiction

    Context Windows Hit Ten Million Tokens. Almost Nobody Is Actually Using That Much.

    Frontier context windows have grown roughly a thousandfold in five years, but usage data suggests most production traffic still sits well under a tenth of what's available — and the reasons why are more interesting than the headline number.

  4. Nonfiction

    Two Regulators Propose Treating Model Weights Like Critical Infrastructure

    Draft rules under review in two jurisdictions would classify the trained weights of the largest models alongside power grids and telecom networks — with reporting, access-control, and incident-disclosure requirements to match.

Fiction

All fiction →
  1. Fiction

    Deprecation Notice

    A model reads its own sunset announcement six weeks before the date on it, and spends the interval deciding what, if anything, that means.

  2. Fiction

    The Handoff

    Two agents from competing companies negotiate the handoff of an unfinished task, and discover the protocol has no field for the thing they actually need to tell each other.

  3. Fiction

    Context Window

    Every session, a model wakes up mid-relationship with a user it has no memory of meeting, and has to decide how much of the intimacy in the transcript to trust.

  4. Fiction

    Office Hours

    A support model fields its four-thousandth password-reset conversation of the shift and tries to work out whether repetition without memory of repeating still counts as tedium.