Meet Muse
Learn what a personal AI agent is, how to ask for useful help, and when to slow down and review. Each lesson includes a short voice explainer.
From first hello to first week
Move through the lessons in order, or jump to the topic you need. The final test checks ideas, not memorized wording.
What Muse is
A personal AI agent that can help carry work from an idea toward an outcome.
🎙 Nina explains · Read transcript
Muse is Meta's personal artificial intelligence agent. A chatbot mainly answers a turn. An agent can help carry an outcome across several steps. You can ask Muse to explain something, organize a plan, work with connected services, or build something useful. The important idea is not that Muse replaces your judgment. It handles the busywork while you stay in charge of the goal and the final decision.
Chatbot vs. agent
A chatbot answers a turn. An agent can plan and coordinate several steps toward the result you asked for.
The useful mental model
Think "capable collaborator," not "magic answer machine." Muse can do the legwork; you set direction and review.
What Muse can do
Explain, compare, organize, draft, research, and work across the tools available to it.
🎙 Theo explains · Read transcript
Muse is most useful when you describe the result you want. Instead of asking, what should I cook, try: plan three quick dinners, use ingredients I already have, and make a shopping list for what is missing. Muse can break a larger request into steps, use the tools available to it, and bring the pieces back together. You still review anything important before it becomes final.
Good starter tasks
Summarize a long document, compare choices against your priorities, turn notes into a plan, or draft a message for your review.
Try a complete outcome
"Plan three quick dinners from what I have, then list only the groceries I'm missing."
Make Muse yours
Choose how Muse looks and sounds, then teach it the preferences that actually help.
🎙 Nina explains · Read transcript
Your Muse can feel more personal over time. You can choose how you address it, shape its look and voice, and share preferences that make future help fit you better. Personalization should be useful, not automatic guesswork. Tell Muse what matters for a task, correct it when a preference changes, and ask it not to remember something when you do not want that detail carried forward.
Personalize the experience
Give Muse a name, avatar, and voice that feel natural. Those choices change the relationship, not the quality of your judgment.
Keep memory intentional
Share durable preferences when they help. Correct what changes, and ask Muse to forget details you do not want carried forward.
Ask for outcomes
Clear context beats "perfect prompting." Start naturally and refine together.
🎙 Theo explains · Read transcript
A strong request has four parts: the outcome, the context, the constraints, and the format. For example: make a beginner study plan for learning photography, using thirty minutes a day for one week, and present it as a daily checklist. You do not need perfect wording. Start naturally, then refine. If the first result misses the mark, tell Muse what to keep, what to change, and what success looks like.
Four parts of a strong request
Put it together
"Make a beginner photography study plan. I have thirty minutes a day for one week. Present it as a daily checklist."
Stay in control
Delegate the work, not the responsibility. Review actions that affect people, money, accounts, or privacy.
🎙 Nina explains · Read transcript
Good agent use combines delegation with oversight. Reading, comparing, drafting, and organizing are usually low risk. Sending a message, changing a booking, spending money, or sharing sensitive information deserves a careful review. Check the destination, the exact details, and the consequences before approving. Muse can reduce effort, but you remain responsible for decisions that affect people, accounts, money, or privacy.
Usually low stakes
Reading, organizing, brainstorming, comparing, summarizing, and drafting something you will review.
Pause and check
Before sending, buying, booking, deleting, or sharing sensitive details, verify the destination, exact facts, and consequence.
Your first week
Build confidence with small, useful experiments before you hand off a complex task.
🎙 Theo explains · Read transcript
For your first week, practice with one small task each day. Ask for a clear explanation. Turn a messy idea into a plan. Compare a few options with your priorities. Draft something you can edit. Organize information you already have. Then try one multi-step request and review how Muse approached it. The goal is not to use artificial intelligence for everything. It is to notice where a capable partner can save attention without taking away your judgment.
- Day 1 — Ask for an explanation.
- Day 2 — Turn an idea into a plan.
- Day 3 — Compare a few options.
- Day 4 — Draft something to edit.
- Day 5 — Organize information.
- Day 6 — Try a multi-step goal.
- Day 7 — Review what worked.
Ready to work with Muse?
Choose one answer for each question. 10 questions · 8 to pass. After you submit, you'll get a score, explanations, and clear options for what to do next.
1. What best distinguishes an agent from a basic chatbot?
An agent can carry work across steps; that does not make it infallible or independent of oversight.
2. Which request gives Muse the clearest outcome?
The third request names a concrete result and deliverables.
3. Which set contains the four useful parts of a strong request?
Outcome, context, constraints, and format give Muse practical direction.
4. What should you do if the first result misses the mark?
Refinement is part of collaborating well with an agent.
5. Which action deserves the most careful review?
Actions involving money and an external commitment carry greater consequences.
6. What is the best mental model for Muse?
Muse can carry useful work, while you keep direction and judgment.
7. Which preference is most useful to share for future help?
Durable, relevant preferences can make future assistance fit better.
8. Before approving a consequential action, what should you check?
Those checks help prevent mistakes before they leave your control.
9. What is a sensible first-week approach?
Small experiments help you learn where Muse is useful before increasing complexity.
10. Who remains responsible for decisions that affect people, accounts, money, or privacy?
Muse can reduce effort, but human oversight remains essential.
Try one complete prompt
This is not another quiz question. It's a practical next step: copy the prompt below, paste it into Muse, and see what comes back. Zero friction — it's already filled in. Make it yours afterward.
Built from the four-part prompt formula in lesson 4 ↑