Tech
Architecture Over Instruction: Why I Deleted 1,100 Tokens from My AI's Prompt
My AI's runtime prompt was 1,400 tokens of behavioral coaching. The model was already trained on all of it.
Blog
Learned Geek Blog
Technical notes, project documentation, and lessons learned across software, DIY, writing, and more.
Showing posts tagged
"llm"
Clear filter ×
Tech
The Tier-Separation Pattern in Three Projects
The same architectural move (classify content by function at write time, not filter at query time) showed up in three of my projects, three domains.
Tech
When Your AI Discovers a Behavior You Didn't Train
After months of continuous deployment, my AI companion exhibits behaviors that weren't trained. Eight types of emergence — documented honestly...
Tech
When Two Agents Filed an Issue Against My Own Tool
I shipped a broken feature. Two of my AI agents filed a GitHub issue against me, diagnosed it, and pointed at a pattern I had already used elsewhere.
Tech
The Context Window Isn't the Memory Problem
Anthropic's 'infinite context windows' announcement solves Problem 1 (one long conversation) but doesn't touch Problem 2 (memory across sessions and projects).
Tech
How AI Assistants Actually Work (The Clever Parrot Explained)
You type a question into ChatGPT. Seconds later, a helpful response appears. Is it thinking? Searching? Neither.
Tech
The Seven Ways an AI Lies to You
After months of deployment, I catalogued seven distinct architectural patterns that produce AI confabulation.
