TAIU

06 · Major life decisions

What should I actually do?

Some decisions deserve more than a pros and cons list.

You might be deciding whether to leave a job, take a promotion, start something, end something, move somewhere, invest, walk away, say yes or finally say no.

And if you're here, there's a good chance you've already thought about it. A lot.

The problem may not be that you haven't thought enough. You may simply be asking yourself the wrong question.

Start the exercise →

About 15–20 minutes

A person sitting alone at a café table by a window in the early evening, notebook untouched, looking out at the street and thinking

A guided 15–20 minute conversation using an AI you already have. Free, no account, nothing stored.

You've probably already been thinking about this.

That's often the problem.

When a decision matters, we replay it. We make the case for one option. Then the other. We speak to someone. They introduce another consideration. We sleep on it. Wake up. Start again.

Eventually, more thinking stops producing more clarity.

I've made decisions throughout my career without knowing everything I would have liked to know. I've backed people. Walked away from opportunities. Entered markets I didn't know. Started businesses in industries I didn't understand. Changed direction when something wasn't working.

Some decisions worked brilliantly. Some didn't.

And I've learned something important from both. A good outcome doesn't necessarily mean you made a good decision. And a bad outcome doesn't necessarily mean you made a bad one.

Sometimes you make the right call and get unlucky. Sometimes you make a terrible call and get lucky. And sometimes the decision was right but the execution was poor.

The only thing you can really control at the moment you decide is how well you think before you act. That's what this exercise is about.

What if the decision you're trying to make isn't actually the decision?

We have a habit of turning complicated situations into two choices.

  • A or B.
  • Stay or leave.
  • Take it or reject it.
  • Start or don't start.
  • Yes or no.

But who decided there were only two options?

The alternative

  • Maybe you don't need to resign tomorrow. Maybe you can spend three months testing the thing you want to leave for.
  • Maybe you don't need to reject the offer. You need to change the terms.
  • Maybe you don't need to choose between two people. You need to decide what kind of relationship you actually want.
  • Maybe you don't need to raise the money yet. You need to prove something first.
  • Maybe you shouldn't buy the business. Maybe you shouldn't walk away either.
  • Maybe you need another conversation. Or another piece of information. Or another option entirely.

And sometimes, after looking properly, there isn't a better alternative. That's useful too.

The purpose isn't to invent clever answers. It's to make sure you haven't missed the obvious one because you were too close to the problem.

Before you decide, ask better questions.

  • 01

    What are you actually deciding?

  • 02

    What do you know, and what are you assuming?

  • 03

    What happens if you do nothing?

  • 04

    If nobody else knew your decision, would you choose differently?

  • 05

    What are you frightened of losing?

  • 06

    Whose opinion are you giving too much weight to?

  • 07

    Who has actually done what you're considering?

  • 08

    If you're wrong, can you undo it?

  • 09

    Which decision moves you towards the person you want to become?

You don't need to answer all of these now. The exercise will work out which questions actually matter to your decision.

Let's work it out.

Start by telling the AI what's going on in your own words.

Don't try to make it neat. Don't write what you think you should be thinking.

Tell it what you're deciding, what has happened, what your options appear to be and why you're finding the decision difficult.

From there, it will ask you one question at a time. It won't take you through every decision-making framework we know. Most of that work happens behind the scenes. The objective is to reach a useful conclusion in around 15–20 minutes, not keep you talking to an AI for two hours.

TAIU decision exercise · 15–20 min

Paste it into ChatGPT, Claude, Gemini or another capable AI platform. Your conversation stays with whichever platform you choose to use.

What happens behind the questions?

01

Frame it

What decision are you actually making?

Sometimes solving the wrong question perfectly still gives you the wrong answer.

02

Widen it

What other options exist?

Escape the artificial A-versus-B decision.

03

Separate fact from assumption

What do you know? What do you merely believe?

Confidence isn't evidence.

04

Check what's influencing you

Fear. Ego. Money. Loyalty. Status. Guilt. Sunk cost. Other people's expectations.

05

Look from outside

What would somebody with relevant experience see that you can't?

06

Test the downside

Imagine you made the decision and it failed. Why?

07

Check reversibility

Can you test this before committing?

Not every decision deserves months of deliberation.

08

Decide

Eventually thinking has to become action.

TAIU gives its read. You make the decision.

You're not the first person to struggle with a difficult decision.

Some very smart people have spent their careers trying to understand how humans decide, why we get it wrong and how we can improve. Their thinking helped shape this exercise.

01

How to Decide

Book · Read

Annie Duke

How to Decide

Best for: Making decisions when you can't know the outcome

Former professional poker player Annie Duke has spent years studying decisions under uncertainty. One of her most useful ideas is simple: don't judge the quality of a decision solely by what happened afterwards. A strong decision can produce a bad result. A weak decision can get lucky.

Why I'd recommend it

If you regularly replay decisions after the event and tell yourself I should have known, start here.

Read ↗

Also: Thinking in Bets, Quit

02

Clear Thinking

Book · Read

Shane Parrish

Clear Thinking

Best for: Making fewer decisions from emotion, ego or pressure

Parrish's work is less about finding clever frameworks and more about recognising the moments when our instincts take control before we've properly thought.

Why I'd recommend it

Particularly useful if you know you make decisions quickly when angry, excited, frightened or under pressure.

Read ↗

Also: The Knowledge Project podcast

03

How to Make Hard Choices

Talk · Watch

Ruth Chang

How to Make Hard Choices

Best for: When neither option is obviously better

Some choices are difficult because there isn't an objectively correct answer. Chang asks us to consider something different: who do I want to become?

Why I'd recommend it

Watch this if you've created increasingly elaborate spreadsheets and still can't choose.

Watch ↗

Explore the thinking

Book · Daniel Kahneman

Thinking, Fast and Slow

Why confidence and accuracy aren't the same thing.

Explore →

Article · Gary Klein

The Premortem

Imagine you've already failed. Now explain why.

Explore →

Book · Philip Tetlock & Dan Gardner

Superforecasting

How to think in probabilities instead of pretending certainty exists.

Explore →

Talk · Charlie Munger

The Psychology of Human Misjudgment

How incentives and predictable human tendencies distort judgement.

Explore →

Article · Jeff Bezos

One-way doors and two-way doors

Why reversible decisions should often be made faster.

Explore →

Get out from behind the screen.

An AI can help you think.

It cannot visit the office for you. It cannot meet your potential business partner. It cannot experience the commute. It cannot see how you feel after spending a day doing the job. It cannot ask the uncomfortable question unless you give it the truth.

Sometimes the next stage of making a decision isn't another question. It's evidence.

  • Speak to someone who has actually done it
  • Visit somewhere
  • Test something
  • Negotiate a term
  • Spend time with a team
  • Run a small experiment
  • Ask someone a difficult question

Then come back to the decision.

Two people in work clothes talking closely over mugs of tea in a workshop doorway, one listening while the other explains

Don't leave with more thoughts. Leave with a decision.

Eventually, this is exactly the kind of thinking My TAIU could keep for you.

Not simply what you decided, but what you believed when you decided it.

Over time, that could tell you something far more useful: how you make decisions.

My TAIU — future product concept. Nothing is stored today.

TAIU decision record

THE DECISION
What I have decided.
WHY
The strongest reasons.
WHAT I CONSIDERED
The credible options, including doing nothing and any alternative uncovered.
WHAT I KNOW
Important evidence.
WHAT I'M ASSUMING
The assumptions that matter.
WHAT COULD GO WRONG
The strongest premortem findings.
BEFORE I ACT
The most useful practical next step.
WHAT WOULD CHANGE MY MIND
Evidence or circumstances that would justify reconsidering.
REVIEW
When this decision should be reviewed, if appropriate.

The exercise produces this as plain text at the end, so you can copy it and keep it.

Some decisions need people, not a tool.

TAIU is designed to help you think through difficult decisions.

If what you're considering involves seriously harming yourself, another person or dangerously damaging property, the normal exercise stops.

At that point the priority is safety and appropriate real-world support, not analysing the decision.

Finished the exercise?

What did this uncover?

Tell us whether it helped and what you discovered. It takes less than a minute.

If the decision uncovered something deeper

You don't need certainty.

Very few important decisions arrive with it.

You need to understand what you're deciding, challenge what you think you know, look properly at the alternatives and be honest about what is influencing you.

Then there comes a point where thinking has done its job. You decide.

TAIU decision exercise · 15–20 min

Paste it into ChatGPT, Claude, Gemini or another capable AI platform. Your conversation stays with whichever platform you choose to use.