Learn AI Roadmap

How It Works

Share what you already know (or don't), how many hours per week you can actually commit, and the concrete outcome you want in 3–6 months. The roadmap produces a 4-phase plan that starts from fundamentals or your actual starting point, moves through prompting and tool use, then projects and specialization. Every phase has 1–2 portfolio outputs, recommended free or low-cost resources, and a simple way to test you are ready before advancing.

What to Provide

Learn AI Roadmap Blueprint

This is the finished deliverable. Fill the [[tokens]] and treat the phases as your actual calendar.

1. Skill-Gap Assessment

Starting point: [[e.g. "I use ChatGPT casually for writing and can do basic Excel. No Python or ML background."]]

Target outcome in 16 weeks: [[e.g. "Ship a personal AI side project that saves me 4 hours/week at work and can be shown in a portfolio."]]

Biggest gaps to close:
- Prompt craft beyond basic chat
- Tool selection and chaining (which model when)
- Turning ideas into small working tools
- Documenting and presenting what you built

2. 4-Phase Plan (16-Week Example at 6 hrs/wk)

Phase 1: Foundations & Prompting (Weeks 1–4, ~24 hrs total)

Goal: Stop guessing. Get consistent good outputs.

• Core concepts: tokens, context window, temperature, system vs user prompts, few-shot.
• Daily practice: 20 prompts a day using the 5-part formula (role, goal, context, constraints, format).
• Resources: LearnPrompting.org free course (first 6 modules), one 30-min YouTube deep dive on context engineering per week.
• Checkpoint project: Build your personal "prompt library" Notion or doc with 15 reusable prompts for your real life/work. Test each 3 times and note what broke.

Phase 2: Tools & Workflows (Weeks 5–8)

Goal: Know which tool for which job and how to chain them.

• Image, video, audio, code, research, spreadsheet, agent tools.
• Build 3 small automations: one research brief, one content repurposer, one data cleaner.
• Resources: Official docs + 1 short course per major tool (e.g. Cursor for code, Midjourney docs, Zapier/Make basics).
• Checkpoint: One 3-tool workflow that produces a weekly deliverable you actually use.

Phase 3: Projects & Specialization (Weeks 9–12)

Goal: Ship real things and pick your lane.

• Choose 1 focus area (writing + research, image systems, code tools, personal automation, teaching).
• Ship 2 portfolio projects that solve a real problem for a real user (even if that user is future-you).
• Document process, prompts, failures, and final output.
• Resources: Domain-specific tutorials + "build in public" threads from people 6 months ahead of you.

Phase 4: Polish, Portfolio, Next Level (Weeks 13–16)

Goal: Make the work visible and decide what is next.

• Refine 2 best projects into case studies (problem → approach → result + before/after).
• Create a simple site or Notion page or PDF "AI Work Samples".
• Optional: apply learnings at work for 2 weeks and measure time saved or quality lift.
• Decide: job hunt, freelance offers, internal promotion case, or bigger personal project.

3. Weekly Time Box Example (6 hrs/wk)

• Mon 1.5h: New concept or tool tutorial + notes
• Wed 1.5h: Prompt / workflow practice + iteration
• Fri 1.5h: Build or document the project piece
• Sun 1.5h: Review, fix, publish small update, plan next week

Protect the time like a meeting with your future self. Missing a week? Do not double the next week. Just resume.

4. Recommended Resources (Current as of 2026)

Free / high signal:
- LearnPrompting.org
- Official model docs (OpenAI, Anthropic, Google)
- Hugging Face courses (free)
- Cursor / Windsurf / VS Code + Copilot docs and YouTube channels

Low cost paid (pick 1–2):
- One focused course on your main tool
- Midjourney or Leonardo subscription for image work
- A good LLM plan with higher limits

Avoid buying 5 courses at once. Finish one project with free resources before paying for the next.

5. Success Checkpoints (Do Not Skip)

• End of Phase 1: Can you take a messy request and turn it into a prompt that gives usable output on the first or second try 80% of the time?
• End of Phase 2: Do you have at least one workflow you run every week that saves you real time?
• End of Phase 3: Do you have 2 portfolio pieces you are proud to show and can explain in 2 minutes?
• End of Phase 4: Can you articulate what you are good at now and what you want to get great at next?

6. Portfolio Output Examples by Goal

• Job switch into AI-adjacent role: 2 case studies + a 1-page "How I use AI to [[your old skill]] 3x faster" summary.
• Automate your job: Before/after time logs + the actual tool or prompt chain + 1 slide deck you used to get buy-in.
• Teach or content: 4 finished pieces (essay, video script, course outline, newsletter) created with your system + the prompts/templates you now sell or share.

Format Checklist

Follow the phases in order. Do not jump to "build agents" in week 2. The people who ship are the ones who do the boring fundamentals first and actually finish small projects.

7. Motivation & Sustainability Notes

• Track "prompts that worked" and "time saved this week" in a simple log. Seeing the wins keeps you going.
• When you feel stuck, go back to Phase 1 prompting practice for 30 minutes. It almost always unstuck the next project.
• Find 1–2 people 3–6 months ahead of you and read what they ship publicly. Copy their process more than their tools.

Do the work in public if it helps you stay accountable. A short weekly update thread or private note to a friend is enough.

8. Quick Start This Week (Even If You Only Have 3 Hours)

1. Spend 45 min on LearnPrompting.org module 1–2.
2. Write 5 prompts for things you actually do this week using the formula.
3. Pick the one that felt worst and iterate it 3 times.
4. Save the best version. That is your first library entry.

Small consistent reps beat heroic weekend binges.

9. Example 16-Week Calendar Snapshot

Week 1–2: Prompt formula + 30 prompts logged. Library started.
Week 3–4: First automation (research brief). Test on 3 real tasks.
Week 5–6: Image or content tool deep dive. 2 portfolio pieces started.
Week 7–8: Combine tools into one weekly workflow you ship every Sunday.
Week 9–12: Main specialization project. Document failures publicly or privately.
Week 13–14: Second project or polish first. Case study draft.
Week 15: Portfolio page live. 1 week of real-world use + measurement.
Week 16: Review, decide next 90-day focus, celebrate the reps.

Print this or put it in your calendar with actual dates.

(150+ line depth for aihowto.pro Learn AI Roadmap generator. [[tokens]] for all personal inputs.)