The Self-Educator
AI as advisor, librarian, tutor, editor, and explorer — on your own machine, at your own pace.
This is not a course. It is a system. Five roles that an AI agent can fill — each one replacing a function that used to require another person or an expensive institution. You build it once, on your own stack, and it works for you across every subject you want to learn.
1. Academic Advisor
Tell the AI what you want to learn — a skill, a subject, a craft. It builds a curriculum designed around you: your current level, your available time, your learning style. It sequences topics, recommends resources, and adjusts as you progress.
No rigid course. No semester schedule. Just a path that fits you.
2. Librarian
Give it a topic. It searches across models and sources — articles, papers, books, discussions — and returns only the handful worth your time. It summarizes each one and tells you why it matters.
Thousands of results become five sources. The needle, not the haystack.
3. Midnight Tutor
Explain what you think you know. The AI finds the exact gap in your understanding — the point where your explanation breaks down — and teaches that specific gap. No judgment, no waiting, no skipping ahead.
It teaches what you don't know, not what you already understand.
4. Editor
Send it your writing, your argument, your draft. It pressure-tests your work instead of praising it. It finds weak logic, missing evidence, unclear phrasing, unsupported claims. Real feedback, not empty encouragement.
It makes your work better because it tells you what's wrong.
5. Explorer Roommate
Tell it what you're interested in. It introduces you to adjacent fields, unfamiliar ideas, and perspectives you wouldn't have found on your own. It builds connections between what you know and what you haven't yet encountered.
It drags you into worlds you'd never explore on your own.
How It Works
These five roles are not separate products. They are the same AI stack — DeepSeek, OpenRouter, DSH — applied in different ways. You build the stack once, then use it in each role.
The DIY AI: Own Your Stack series walks through building each layer. By the end, you have a working self-educator system on your own machine. No subscriptions. No data leaving your control. Just the tools, and what you choose to learn.
Inspired by "How To Become Dangerously Self-Educated With AI" — applied through the DIY AI: Own Your Stack series.
Related: The Stack Explained · Blog post