In the last month or so, I’ve been busy giving presentations on AI-ready content and thought I’d share the “director’s cut,” which includes things I cut out for time in my last presentation at SIGDOC.
In these presentations, I move beyond the basics of prompting to discuss what comes next in our evolving relationship with AI systems.
This presentation captures where my work has been heading over the past year—from the structured prompting techniques we’ve explored in PromptOps to something much more fundamental: how we architect knowledge itself for human-machine collaboration.
The prompt hasn’t disappeared—it’s expanded to encompass entire knowledge systems.
While casual users may not need sophisticated prompt techniques, the principles that made structured prompts effective haven’t disappeared. They’ve evolved.
As AI systems gain the ability to work with uploaded content and knowledge bases, we’re discovering that these same structured principles apply at a larger scale.
The prompt hasn’t disappeared—it’s expanded to encompass entire knowledge systems.
In the presentation, I walk through this progression: from breaking prompts into reusable blocks, to organizing those blocks into libraries, to ultimately structuring the content that AI systems work with.
This mirrors the journey many of us are taking—moving from prompt engineering to what I’m calling context engineering.
I also demonstrate three different approaches to structuring the same information and the results of some simple testing I did to compare outputs.
When we organize content using microcontent, AI systems retrieve and apply that information far more effectively than when we simply dump unstructured notes into a knowledge base.
This research into AI-ready content is shaping the next iteration of my work. I’m revising my Writing with Machines course around this exact progression, and paid subscribers will have free access to audit the updated material as it develops.
I’m also developing methods for testing whether content is truly AI-ready—practical approaches you can use to evaluate your own knowledge systems. More on that coming soon.
For now, I hope this presentation gives you a broader view of where machine rhetorics is heading.
The future of rhetoric isn’t about choosing between humans and machines. It’s about designing knowledge systems that shape our ways of knowing the world.
