Planning & Deciding

A trip, a wedding, a move, a monthly budget, a career choice — the hard part is rarely the doing, it's the organising and the deciding. AI is a superb thinking partner for both: it turns messy thoughts into structured plans, and it asks the questions you forgot to ask yourself.

What AI planning looks like

Give AI a goal plus your constraints, and it hands back structure: days, steps, checklists, timelines, tables. A few everyday examples of what one message can produce:

Trip itinerary. "4 days in Sylhet, family of 5 including grandparents, moderate budget, we like tea gardens and food more than hiking. Day-by-day plan, with a rough daily cost estimate and a rainy-day alternative for each day."
Event plan. "My parents' 40th anniversary, about 60 guests, at home. Give me: a guest-list template, a 4-week countdown checklist, a day-of timeline, and a task list I can split among my siblings."
Meal plan + grocery list. "One week of dinners, Bangladeshi home cooking, under 45 minutes each, two of the days vegetarian. Then combine everything into one grocery list sorted by market section."
First-draft budget. "My monthly take-home is X. Here's roughly what I spend on rent, food, transport… Build me a budget with a savings target and show me the two easiest places to cut."
Study or work schedule. "I have exams in 6 weeks, these 5 subjects, weakest in the first two. Weekly schedule, 3 hours a day, Fridays lighter."
Moving-house checklist. "Moving flats in 3 weeks with a toddler. Checklist ordered by week, including the boring things people forget — meter readings, address changes, deposits."

Notice the pattern in every one: goal + constraints + desired format. Constraints aren't a burden to the AI — they're fuel. "4 days, 5 people, grandparents, moderate budget" produces a dramatically better plan than "plan a trip to Sylhet". (This is the specificity rule from Prompting 101 doing heavy lifting.)

From messy head to structured plan

The worry loop on the left has no checkboxes. The plan on the right does — and a plan you can see is a plan you can improve.

Plans are drafts — kick the tyres

The first plan is never the final plan, and that's the point. Because it took thirty seconds instead of three evenings, you can afford to revise it ruthlessly:

"Day 3 is too packed — spread it out and drop the least interesting stop."
"Make the whole trip 20% cheaper and tell me exactly what I'm giving up."
"My mother can't walk long distances — adjust every day for that."
"Turn the checklist into a table with columns: task, owner, deadline, done."

Treat the plan like clay, not stone. The revision conversation usually adds more value than the first draft — the same iterate-don't-restart habit from Iterating.

Try it now

Plan a [number]-day trip to [place] for [who's going].
Budget: [rough total or "moderate"]. We care most about: [food / nature / history / rest].
Constraints: [e.g. one traveller is elderly, we travel by train, no early mornings].

Give me: a day-by-day itinerary, rough daily cost, one rainy-day alternative per day, and a packing list. Then ask me 3 questions that would make this plan better.

Deciding: AI as sparring partner

Planning organises the work; deciding is harder — it's about trade-offs. Here AI shines not by giving you the answer, but by improving the quality of your thinking. Five tools worth memorising:

Move	What you say	What it gives you
Pros & cons	"List the pros and cons of taking this job offer, given what I've told you."	A balanced starting map — often including angles you hadn't considered.
Better questions	"What questions should I be asking before I decide this?"	The single most underrated prompt in this book. It reveals the blind spots.
Comparison table	"Compare these 3 flats in a table: rent, commute, size, light, landlord flexibility."	Options lined up side by side, so your gut has something honest to react to.
Devil's advocate	"I've decided to open the shop. Argue against my plan as hard as you can."	The strongest case for the other side — before reality makes it for you.
Pre-mortem	"Assume this plan failed badly one year from now. Write the story of why."	Failure modes while they're still cheap to fix. Planners' favourite trick.

AI is naturally agreeable — left alone, it tends to cheer your plan on. The devil's advocate and pre-mortem prompts exist precisely to switch that off. Use them for any decision that costs real money or real years.

Try it now

I'm choosing between these options: [option A], [option B], [option C].
What matters to me, most important first: [e.g. monthly cost, time with family, long-term security].

1. Build a comparison table scoring each option against each factor.
2. Tell me what important factor I've probably forgotten.
3. Ask me the 3 questions whose answers would most change the result.
Don't tell me which to pick — help me see clearly.

Try it now

Here is my plan: [describe your plan in a few lines].

Run a pre-mortem: it's one year later and this plan has failed badly. Write the 5 most likely reasons why, ranked by probability. For each, tell me one cheap thing I could do THIS WEEK to reduce that risk.

The rule: it structures, you decide

Keep this division of labour straight and AI planning is nearly all upside:

AI doesn't know your life. It doesn't know that your sister-in-law will be offended by seat placement, that day 2 is also your anniversary, or that "moderate budget" means something very specific to your bank account. Feed it those facts or expect a generic plan.
It doesn't know your money. A budget it drafts is arithmetic on the numbers you gave — sensible structure, zero knowledge of your real obligations. And for investments specifically: it's a fine explainer of concepts, but it is not a licensed adviser, and big money decisions deserve a qualified human.
It doesn't know local prices or today's reality. Hotel rates, bus schedules, ticket prices, whether that restaurant still exists — its built-in knowledge may be outdated or simply invented. Verify anything local, priced, or bookable before committing (tools with web search help, but check the sources — see Hallucinations).

A brilliant consultant who's never been to your town

Imagine a sharp consultant flown in from abroad: superb at structure, checklists, and hard questions — but they've never walked your streets, met your family, or seen your bank balance. You'd happily let them organise your thinking. You'd never let them sign the contract. That's exactly the right relationship with AI: it structures, you decide.

The "ask me questions first" trick

For any big plan, start with: "Before you plan anything, ask me the 5 most important questions." The AI interviews you, you answer in one message, and the resulting plan fits your life instead of the average life. Thirty extra seconds, twice the quality.

Key takeaways

Goal + constraints + format = a structured plan in seconds: trips, events, meals, budgets, schedules, moves.
The first plan is clay, not stone. "Day 3 is too packed", "make it cheaper" — revising is where the value is.
For decisions: pros/cons, "what should I be asking?", comparison tables, devil's advocate, pre-mortem.
AI is agreeable by default — explicitly ask it to argue against you before you spend real money or years.
It doesn't know your life, your money, or local prices. Verify anything local, priced, or bookable.
The rule that keeps it all safe: AI structures your thinking; the decision — and its consequences — are yours.