My thinking behind this Blog Site.

I don’t write anymore. Not in the traditional sense.

I don’t sit with a blank page, wrestling sentences into shape. I don’t write code either—not a single line. What I do, all day, is type into a small chat box. Sometimes I just talk to it.

And yet, here’s a blog. Full of writing. Full of technical ideas. How does that work?

These posts are AI-assisted, but they’re not AI-generated content. That distinction matters.

I don’t ask an AI to brainstorm blog topics and then generate posts about them. That’s slop—low-effort, hollow, SEO-bait churned out at scale. I understand why people are skeptical of AI content. A lot of it deserves that skepticism.

What happens here is different. I spend hours working on personal projects—building agents, experimenting with frameworks, debugging weird edge cases. I do this work with an AI, talking through problems, trying solutions, hitting walls, finding workarounds. That’s the real work.

The blog posts are artifacts of those sessions.

When I figure out something useful—how to track token usage per module in DSPy, how to manage tools in an agent, how to stream responses in a specific way—I ask the AI to help me turn that session into a post. The knowledge is real. The problems were real. The solutions emerged from actual work.

The word elicited captures this. To elicit is to draw out, to bring forth what wouldn’t otherwise surface. These posts couldn’t exist without the technical work that produced them. They’re not manufactured content; they’re documentation of genuine exploration.

If you’re reading this, you’re probably curious about what that looks like. The posts here reflect what I’m actually building and learning. Nothing more, nothing less.

—S