Human Judgement in the Age of AI.
You already know what to do. You just haven’t closed it.
Decision Closure for people carrying a choice that should have been settled months ago.
You had enough information six weeks ago.
You have more now. You are not closer.
You have run the options. You have asked two people you trust. You have asked AI to check your reasoning, and it produced a careful answer, three considerations you hadn’t weighed, and a slightly different conclusion than the last time you asked.
So you are still holding it.
The cost is not the decision. It is the holding.
AI has made thinking almost free. It has not made stopping any easier.
A machine can generate for as long as you keep asking. It does not feel the weight of a decision left open. You do.
You carry it into meetings. You carry it to bed. You lose the opportunity while you deliberate. You pay for the delay in a currency nobody itemises — attention, momentum, energy that should be going somewhere else.
And every fresh round of analysis feels like diligence, right up until you notice it is the thing keeping the decision open.
This is not indecisiveness.
You decide well elsewhere. You do it daily, at speed, and you’re usually right.
What has changed is that the number of defensible options in front of you has grown faster than your confidence in choosing between them. AI can build a persuasive case for almost any direction. That used to be the useful part. Now it means the weak options no longer eliminate themselves.
More rational possibilities do not produce one responsible decision.
Someone still has to close it.
You may recognise some of this:
- You keep going backwards and forwards over the same ground.
- You settled it, and then you reopened it, and you are not sure what new evidence justified that.
- You cannot tell whether you are being careful or simply afraid.
- You do not know how much more research is enough.
- You have a recommendation you could defend, but you would rather not have to.
- The cost of waiting has quietly become larger than the cost of being wrong.
Different decisions. Same mechanism.
Whether it is a business you are deciding whether to buy, a role you are deciding whether to leave, an architecture you have re-evaluated four times, or a conversation you keep not having — the structure underneath is identical. A decision that is open, expensive, and being kept open by the belief that more thinking will settle it.
What closure actually is.
Not certainty. Certainty is not available and waiting for it is the trap.
Closure means the decision stops occupying you.
It means you know what you decided, on what basis, what you accepted as the cost, and what specific evidence — not what discomfort — would justify opening it again.
That is a different thing from being sure. It is more useful, and unlike certainty, it is achievable this week.
Hello, I’m Manoj,
I have made decisions about businesses to pursue, markets to enter, acquisitions to reject and directions to leave behind. Some were straightforward. Several were not, and a few cost me real money and years while I held them open.
I have used AI daily since January 2023 to research, challenge and organise those decisions. It has been genuinely useful. It has also shown me its limit clearly: it can build the case, and it can build the counter-case, but it cannot decide what I am prepared to accept or carry the consequence for what follows.
That gap — between assistance and authority — is where this work sits.
Where to go next.
If you want the full argument — what human-led actually means, where AI’s authority should end, and why oversight so often becomes ceremony:
If you would rather sit with this a while first, I write one letter a week on the same territory. Free, and you can leave whenever you like.
~Manoj
Human Judgement In The Age of AI
Clear. Capable. Human.
