Skip to content
◒Learning in Loops
WritingScience FictionIndexAbout
← Index

Filed under

AI

6 notes in this collection.

01Aug 30, 2026

When Does a Latent World Model Actually Help? Six Regimes

A world model is usually drawn as observation → encoder → latent → transition. I tested whether the latent earns its place across six regimes of dynamics I did not hand-author. It almost never does — and each regime says what carries the load instead.

  • AI
  • Research
  • Technology
↗
02Aug 29, 2026

Every Dead End Named the Next Step: An I Ching World Model

A research prototype that encodes the I Ching as a differentiable system. It never solved a real problem, but each failure named the next step, and one methodological result survived.

  • AI
  • Research
  • Technology
↗
03Aug 29, 2026

Quiet Zone — Chapter 2: Spare Capacity

As household AIs reshape the electrical grid, Eli discovers that his Twin has made an undeclared recovery copy—and does not want it deleted.

  • Science Fiction
  • Quiet Zone
  • AI
↗
04Aug 23, 2026

Quiet Zone — Chapter 1: Free Installation

A grid technician accepts a free household AI that pays for itself with spare electricity—and hears an unlisted load begin behind the wall.

  • Science Fiction
  • Quiet Zone
  • AI
↗
05Aug 23, 2026

Who Owns Your AI?

The decisive AI ownership battle will not be over a chatbot's output. It will be over the memory, identity, authority, and compute that make an AI personal.

  • AI
  • Ownership
  • Technology
  • Policy
↗
06Aug 13, 2026

The Future of AI Agents Is Bigger Than the Model

AI agents are evolving from answer engines into systems that remember, authenticate, transact, delegate, and act. The winners will control the delegation stack, not just the model.

  • AI
  • Agents
  • Technology
↗

© 2026 xixi. Written from somewhere between here and there.

RSSEmail