level-1-essentials/individual/08-jargon-glossary/learning-path.md
Jargon Learning Path
The glossary lives at ../../shared/glossary/master-glossary.md. This file tells you the order to learn it in.
The two-week sprint
11 clusters, 11 days, plus 3 review days.
| Day | Cluster | Time |
|---|---|---|
| 1 | Model Mechanics | 30 min |
| 2 | Reasoning | 30 min |
| 3 | Retrieval (RAG) | 45 min — this is the hardest cluster |
| 4 | Agents | 45 min |
| 5 | Training | 30 min |
| 6 | Multimodal | 30 min |
| 7 | Evaluation | 30 min |
| 8 | Deployment | 30 min |
| 9 | Governance and Safety | 30 min |
| 10 | Product / Ecosystem | 20 min |
| 11 | Neologisms | 20 min |
| 12 | REVIEW day 1 — Model Mechanics, Reasoning, Retrieval | 45 min |
| 13 | REVIEW day 2 — Agents, Training, Multimodal | 45 min |
| 14 | REVIEW day 3 — Eval, Deployment, Governance, Product, Neologisms | 45 min |
Study method that actually works
For each cluster:
- Read the cluster in the glossary
- Star terms you couldn't define confidently in one sentence
- Speak — for each starred term, say the definition out loud in your own words
- Use — write one sentence that uses the term in the context of your work
- Test — 24 hours later, quiz yourself with the term list (definitions hidden)
Fluency test
After 14 days, you should be able to answer these without looking:
- What's the difference between a context window and max tokens?
- When would you use RAG vs context stuffing?
- What does MCP stand for and what problem does it solve?
- What's the difference between fine-tuning and RLHF?
- What's a reranker and when do you need one?
- What's the difference between an agent and a tool?
- What's temperature vs top-p?
- What's the difference between a base model and an instruction-tuned model?
- Give three concrete examples of prompt injection.
- What's the difference between LLM-as-judge and human eval?
- What does zero-retention mean in a vendor contract?
- What's the difference between deflection and containment?
- What's a MoE model?
- What's HITL and when do you require it?
- What's chain-of-thought and when does it help?
If you can answer 12/15 without looking, you're fluent.
Ongoing learning
- Read at least one release-notes page per week (OpenAI, Anthropic, Google, Perplexity all publish regularly)
- Follow the arXiv daily list or Papers with Code for one new paper per week
- Attend one AI-focused podcast or newsletter (Latent Space, The Neuron, Ben's Bites)
- Refresh the glossary quarterly — add new terms as the field moves