Layered writing pages, notes, and light on a dark desk.

Writing

Essays, models, and working notes.

Sideways Thought thinks in public about artificial intelligence, learning technology, accessibility, institutions, service design, and the behavior of systems once people start using them.

Newest pieces first. Some are finished essays; others are working models, arguments, or notes that may become something larger.

  • Agents Have Needs Too! How to Successfully Onboard Agents in Your Workplace

    Agents Have Needs Too! How to Successfully Onboard Agents in Your Workplace

    Introduction AI agents are becoming more capable, more persistent, and more useful in day-to-day work. Afterall, OpenClaw is to-date one of the most popular agent framework, and open source projects in the world. But OpenClaw, Claude Code, CODEX and other services are not plug-and-play. If an organization wants agents to produce reliable work and reliably…

    Read more

  • Speaking Machine: A Language to Communicate Facts to Large Language Models

    Speaking Machine: A Language to Communicate Facts to Large Language Models

    The way we talk to large language models needs to change. Fact-RAR is an attempt at describing information to Large Language Models that takes advantage of their knowledge of our languages.

    Read more

  • What If Language Models Could Feel Time? Introducing the Time Token Hypothesis

    What If Language Models Could Feel Time? Introducing the Time Token Hypothesis

    Technology has long struggled to capture the nuance of how people actually move through the world. Our systems exist outside the flow of time. Sure we have calendars, scheduling tools, timers. But all of those features are designed for us, and are calculated with raw math and complex libraries. But our systems can’t “feel” time.…

    Read more

  • AI’s False Balance: The Truth Problem

    AI’s False Balance: The Truth Problem

    ,

    Big tech says balance. But when AI treats lies and facts as equal, who loses? The Key Question When Meta announced that its LLaMA 4 model would strive for “balance”—presenting “both sides” of controversial issues—I felt the familiar churn of frustration. The word sounds noble, even responsible. But in practice, this kind of framing often…

    Read more


Browse by thread

Most current writing is about AI, with some classroom and institutional threads emerging around it.

If a piece points toward a practical problem in AI adoption, learning technology, accessibility, or web strategy, there may be consulting work behind it.