https://www.gravatar.com/avatar/a7cc4843f7c957afe8575b1f8c45ef44?s=240&d=mp

Multifarious musings from my meandering mind...

To see posts in Esperanto or Shavian (𐑖𐑱𐑝𐑾𐑯), use the language selector at the top of the screen.

My AI Teaching Setup, Part 0: A Life in Plain Text

This is the first of a few articles about the AI system I use for teaching maths. Basically, ten months ago I decided to buy an AI subscription to learn more about the technology, and today I run a small fleet of machines and agents that draft my lessons, check my exams, index my syllabus, and argue with me about design at midnight. This first article is about the recent years before any of that, because it turns out to be the part that made everything else possible.

My AI Teaching Setup, Part 1: Rented Intelligence

Part 0 was about the plain-text foundation. This is the part where the subscriptions arrived β€” and the part where I should confess that the whole edifice started, like most things worth doing, with fun.

January: renting intelligence

In January this year I bought two AI subscriptions: Claude, and Google’s AI Pro tier. The teaching use case was immediate and real. Claude drafted lesson structures, rewrote exam questions to shift difficulty without shifting marks, generated first-draft assessment tasks from a paragraph of requirements, and proofread things I had read too many times to see. A set of lesson notes that used to start from a blank page now started from a draft I argued with. Arguing with a draft is much faster than producing one.

My AI Teaching Setup, Part 2: The Fleet

Part 1 ended with the 8 GB ceiling and a decision. This is the hardware part, and the part where the system stopped being a computer and started being a staff.

The gaming PC that never gamed

In July I built hyperion: a proper gaming PC, RTX 5090, 32 GB of VRAM. The justifications were two, and I was honest about both at the time. The AI justification: if I was going further into this, I needed headroom β€” enough VRAM for a genuinely good open model with a long context window. The hobby justification: I have wanted a big gaming PC for roughly a decade, and AI was the first excuse strong enough to survive my own accounting.

My AI Teaching Setup, Part 3: The Organs

Part 2 was about models and harnesses. But a model in a harness is still just a clever voice. What turned the setup into a system was the organs: the memory, the senses, and the reflexes bolted on around it. This is the final part, and it starts with a full circle.

The knowledge base that was already ten years old

People ask how I got my teaching knowledge into the AI. The answer is anticlimactic: I didn’t. I had spent a decade writing it in plain text β€” Org-mode notes, syllabus annotations, lesson reflections, the whole zettelkasten from Part 0 .

AI Agents and Local LLMs

In my last article on AI agents I talked about using Claude Code AI agent. This is an easy-to-set-up commercial solution that gives excellent results. The risk is that the pricing may not remain as affordable as it is today, and as a result I wanted to find an alternative solution that insulates me from changing commercials.

There are a number of open-weight models that one may download and run locally on your home computer. Some of the benefits are:

AI Agents Take Over Lesson Prep

It has been a while since I have written about how I use Emacs/LaTeX to generate lesson notes for my students. Quite a lot has happened during that time. One of the main game changers has been the advancements in “AI” technology. Over the next few articles I intend to describe my journey as I navigate this technology and apply it to my teaching practice.

Chat Bots

Having decided to dip my toe in the water, I signed up to two of the prevalent players in this space, being Google Gemini and Anthropic Claude. My early experiments with ChatGPT had led to some disappointing results, where it emphatically tried to convince me that the value of Euler’s constant is 1.5. Since then, the technology has advanced significantly, and I thought it worthwhile to try again.