# LLMs: A Primer Presentation for Our Newest, Scariest Tool

> An engineer-focused primer on the mechanisms that drive Large Language Models. Basics on how we got here, how they work, and how to use them without feeling an apocalyptic dread.

- Page: https://blog.chaoticgood.computer/content/annotations/llm-starter
- Author: Spencer Elkington
- Published: 2025-07-15
- Updated: 2026-11-02
- Tags: engineering/ai, engineering/languages/csharp, engineering/languages/typescript, projects/teaching, writing/essays, writing/annotations

---

This is a presentation I wrote up about LLMs back in early 2025 to try and help other developers understand them mechanically and dispel a some of the doom-and-gloom, fear-of-the-unknown resistance to new tooling I was seeing at the time.

> [!note] Note
>
> If you’re looking to fast-forward into the future, and see my response to “why”, [Software Gardening](https://blog.chaoticgood.computer/content/notes/software-gardening) is the area where I continue this presentation with an extra two years of experience.

Source document: [raw.githubusercontent.com](https://raw.githubusercontent.com/chaoticgoodcomputing/blog/refs/heads/main/content/public/assets/llms.pdf)

## Annotations

> Whydoes the introduction of thistool change how we defineour work?

A comment from *far* in the future. When I wrote this presentation, I stopped here intentionally as a chance for an open-ended discussion about what this meant. This part ended up being *far quieter* than I expected, probably from the sheer scope. Asking this question in June of 2025 would be like asking, in 2008, if a site written by some dweeb in a Harvard dorm would eventually connect you to everybody you’ve ever met or sway the results of an election; it is, in the most literal sense, un-predictable.

Sep 30, 2026