How to Use Muse to Make a Decision

Yes, with a clear division of labor. Muse, Meta's personal AI agent, can gather the facts your decision hinges on — salary benchmarks, costs, contract terms, timelines — using its own browser, then structure the choice: criteria, weights, a matrix, and the honest trade-offs. It remembers your priorities and constraints, so the analysis reflects your actual life, and it works asynchronously while you sleep on it. What it will not do is decide for you, and that is the right shape: use it to kill uncertainty, then make the call yourself.

• A decision matrix is only as honest as its weights. Muse will happily justify whatever you subconsciously already want — ask it to argue the other side too.
• Muse is a new agent and can err; Meta's own launch messaging says it will sometimes fail to finish things. Verify any fact the decision truly turns on.

1. 1

   Lay out every option, honestly

   Name all the real options, including 'change nothing.' Half-listed decisions produce confident nonsense.

2. 2

   Say what you are optimizing for

   Not just 'best' — best for whom, over what time horizon, at what risk tolerance. Muse weights the criteria accordingly.

3. 3

   Send it to fill the factual gaps

   Have Muse research what you don't know — market rates, fine print, timelines — before any scoring happens. Facts first, weights second.

4. 4

   Request the counterargument

   After the matrix, ask Muse to make the strongest case against your preferred option. That is where an assistant earns its keep.

A marketing manager is choosing between a higher-paying corporate offer and a counteroffer from the startup they currently work at.

Goal: Help me decide between a corporate offer ($118k, hybrid, big team) and my current startup's counter ($105k, remote, head of channel) by Friday.
Context: Optimizing for growth first, then pay, then flexibility. Two-year horizon. Medium risk comfort. I have a mortgage and no kids.
Requirements:
- Build criteria from those priorities and tell me if any are missing.
- Research title benchmarks for both roles in my city and attach sources.
- Score both offers plus 'stay without the counter' and show the math.
Actions:
- Look up current salary data for both titles and levels; note where every number comes from.
- Draft a one-page decision memo with a confidence level and the top risk of each path.
Approval: Research only — nothing gets sent to either employer; ask before acting outside research.
Output: Matrix, memo, the strongest case against your recommendation, and the questions only I can answer.

• All the options, stated plainly

  Including the status quo. Decisions go wrong when one option silently never gets weighed.

• Priorities in rank order

  Equal-weighted criteria are how people rationalize. A rank order is what makes the matrix mean something.

• Time horizon

  The best two-year choice and the best ten-year choice are frequently different options.

• Facts versus fears

  Separating what you know from what you suspect tells Muse exactly what to research.

• Your risk tolerance

  It determines how heavily a downside should count, and Muse should apply yours, not a default.

For this task, it is smart to make Muse check with you before it takes certain actions. Decide your rules up front and include them in the prompt.

• Communicating the decision to anyone — Accepting or declining is consequential and personal. Muse checks before sensitive actions like sending an email — this message stays yours.

• Signing, depositing, or committing to anything — The decision memo precedes commitments; it never replaces them.

• Acting on researched facts without verification — If the decision turns on a number, click the source yourself before you move.

• Letting the analysis drift into new tangents — Investigation can delay decisions forever. Tell Muse when the evidence is sufficient and ask for the memo.

• ✓Put 'change nothing' on the list. It is the control group, and it exposes whether you are moving toward something or merely away from something.
• ✓Ask for the math, not just the scores. Seeing which criterion carries the result tells you whether the matrix reflects your priorities or smuggled in someone else's.
• ✓Have Muse research before you weight anything. Most bad decisions are information problems wearing a psychology costume.
• ✓The highest-value question: 'what would have to be true for the other option to win?' It surfaces assumptions you did not know you had made.
• ✓Use the asynchronous gap deliberately: kick off the research, sleep on it, read the memo in the morning. Muse keeps working after you close the app.