← Sovereign AI

⏱ About 15 min✦ 15 XP

Build Your Toolkit

The previous eight lessons have given you a framework. Now it is time to build something real. A sovereign AI user does not just understand concepts in theory — they put that understanding to work by making deliberate, informed decisions about their own tools. This lesson is your workshop session. You will design a personal resilient AI toolkit that reflects your actual tasks, your privacy priorities, and the resilience principles you have studied.

What You Are Building

Your toolkit is a documented plan, not just a list of tool names. A real toolkit design includes: Which tools you will use for which tasks, and why you chose each one. The specific tradeoffs you are accepting — and which ones you are not willing to accept. Your data protection strategy: how you will export regularly, what formats you will use, and where backups will live. Your backup plan for each critical tool. A brief record of what you know about each tool's privacy practices and business model. This documentation serves you in multiple ways. It forces you to think through every decision rather than drifting into habits. It gives you a reference when something changes — you can see exactly what assumptions your toolkit was built on. And it is a living document: as tools change and your needs evolve, you update the plan.

A Toolkit Is a Document, Not Just a Habit

Writing your toolkit down transforms passive habits into active decisions. A documented toolkit can be reviewed, updated, and shared. Unwritten habits are invisible and hard to change.

Before You Begin: Checklist

Before you start designing, review these questions. Your answers will shape your toolkit. What tasks do I want AI help with? (writing, studying, coding, creative work, research, something else?) How sensitive is the information I typically share with AI tools? (low: public topics; medium: personal opinions or plans; high: private health, legal, or financial matters) Do I have hardware capable of running local AI, or am I limited to cloud tools for now? What is my budget? (zero, a small monthly amount, or more?) How much setup and maintenance effort am I willing to invest? Do I work in a context with specific rules about data — a school, organization, or jurisdiction with particular privacy requirements?

The Toolkit Design Framework

Use the following framework as the structure for your design. Fill in each section honestly based on your real situation. Do not design an ideal toolkit for someone else — design a realistic toolkit for you. Section 1: My Use Cases. List each task category where you will use AI. Give each a sensitivity rating (low, medium, or high) based on how private the information involved typically is. Section 2: Tool Selections. For each use case, name your primary tool and a backup tool. Note whether each is open or closed. Note its cost. Note one privacy fact you confirmed from its terms of service. Section 3: Data Strategy. For each primary tool, describe how you will export your data and how often. Name the format it exports in. State whether that format is open or proprietary. Section 4: Resilience Review. For each use case, confirm that you have a backup option. Identify any use case with no backup and name your plan to resolve it. Section 5: One Local Option. Even if you primarily use cloud tools, identify one open-weight model you could run locally and describe one scenario where you would actually use it.

Good Enough Is Real

Your first toolkit does not need to be perfect. A real, imperfect toolkit you have actually thought through beats a theoretical perfect one that exists only as an idea. Start where you are and improve as you learn.

Design Your Resilient Personal AI Toolkit

  1. This is your main project for this lesson. Dedicate at least 20 minutes to completing each section thoroughly.
  2. SECTION 1 — MY USE CASES
  3. List every task category where you currently use or want to use AI assistance. Rate each one Low, Medium, or High sensitivity based on how private the information typically is. Aim for at least three distinct use cases.
  4. SECTION 2 — TOOL SELECTIONS
  5. For each use case from Section 1:
  6. a) Name your primary tool choice and briefly explain why you chose it.
  7. b) Name one backup tool and explain why it qualifies as a real alternative.
  8. c) Label each as open or closed.
  9. d) State the cost of each.
  10. e) Write one privacy fact you found in the tool's actual privacy policy or terms of service.
  11. SECTION 3 — DATA STRATEGY
  12. For each primary tool:
  13. a) Describe how you will export your data and how often you will do so.
  14. b) Name the file format the export uses.
  15. c) State whether that format is open (any tool can read it) or proprietary (only this tool can read it).
  16. d) Describe where your backups will be stored.
  17. SECTION 4 — RESILIENCE REVIEW
  18. For each use case, confirm: If my primary tool disappeared tomorrow, could I continue working with my backup within one day? If the answer is no for any use case, write a concrete plan to close that gap.
  19. SECTION 5 — MY LOCAL OPTION
  20. Research one open-weight AI model you could run locally on hardware you have access to (or that a school or library might have). Write: the model name, where to download it, what software you would use to run it, its approximate hardware requirements, and one specific scenario from your own life where you would actually use it instead of a cloud tool.
  21. FINAL REFLECTION
  22. Write two to three sentences: what is the biggest vulnerability you discovered in your current or planned AI habits, and what one change will you make because of this module?

Why does a well-designed AI toolkit include a documented data export strategy for each primary tool?

A student builds a toolkit with one great cloud AI tool for everything and no backups. Which resilience principle does this violate?

← Back to Sovereign AI