On this page
Here is a question I have been chewing on for a while. If the machine answers in seconds, what is left for us to do? Thinking, and most of us have no time for it.
What mastery of thinking is
Mastery of thinking, as I see it, is three decisions made on purpose. What deserves my head this week. How much time it gets. And what exactly I am asking, before I ask it.
Most of us make all three by accident. The first is the hardest, since most of what lands on my desk only looks urgent. It is also the one I see people trying to hand to AI.
The quality of an answer comes from the clarity of the question. Clarity comes from time spent thinking. There is no shortcut through the machine.
Why it matters now
A poor question always cost a week. The week was also a buffer, and somebody had time to notice the question was poor. Now the poor answer arrives in seconds, and somebody acts on it the same afternoon.
Inside a company the answer does not stay on one screen. It becomes a slide, then a decision, then somebody's work for a month.
Humans drive AI, at least for now. The ones who ask the right and better questions move fastest and get the most out of it.
Ask AI the same question as your competitor and you get two versions of the same plan. It draws on what everybody already knows. Two studies found exactly this. People working with AI did better work one by one, and more similar work as a group (Doshi and Hauser; Anderson, Shah and Kreminski).
What changes the substance is what you put in that nobody else has. Your customers, your numbers, your history, your judgement. Innovation is not born by asking the same questions as everybody else.
Waiting for answers is the one old habit that does not carry over. The thinking habits are the opposite. They matter more now, because nothing else in the process slows you down enough to check.
Who masters it
Everyone should think about this, leaders most. Their questions set the direction for everybody else, machines included. Nobody two levels down improves a question that came from the top. They answer it.
I recognise the ones who master it by three things. They can say in one sentence what they want before they ask. They put their own context into the question, the part the machine cannot know. And they check what comes back instead of forwarding it.
The more confident people are in the tool, the less they think for themselves, a survey of knowledge workers found (Lee and others). I understand why. The answer reads as if somebody had already checked it, and nobody has.
Checking is part of the thinking, and often the biggest part. The answer shows me what my question left out, and the next question is better for it.
How they do it
They know where they are before they plan the route. I have sailed across the Atlantic, and the navigator's first job is to know where the boat is. The boat moves all night, so yesterday's position is already wrong.
A company that plans from last quarter's report plots a route from where it used to be. With AI the wrong plan just arrives faster.
They check the question before it goes out. Before I open my mouth in a meeting, I ask myself: will my comment add any value? Often it will not, so I keep my mouth shut and let others talk. The same question works before I type anything to AI.
What do I want from this, exactly? If I cannot say it in one sentence, I am not ready to ask. The sentence also says who the answer is for. The board and the sales team need two different plans.
Most questions I see also skip what the person already knows, what they have ruled out, and why. That is the part that took thinking. Leave it out, and the answer starts from zero.
Where the time comes from
In the companies I have worked in, too much of a leader's week went to reporting, and to sales reports that were out of date when printed. Your week may look different. Check it.
Why still do it, when the information could be almost live? Habit, mostly. Nobody stopped to ask whether a human still has to do it.
I think most leadership teams would decide better with half the reports and one hour of real thinking a week. Few have that hour in the calendar. The reporting is in it every week.
Before the next report, ask who reads it and what they decide with it. If the answer is nobody and nothing, the report can go. If it is one person and one decision, one page is enough.
What helps a leader
Before any tool, the basics. Stress, poor sleep and being out of shape switch thinking off first. No tool I know of switches it back on, and I have looked.
Then the tools. There are many, and Company Brain is one of them. When they work, the gathering and the reporting go to the machine. That is the promise, and one of these tools is mine, so read this paragraph accordingly.
In the worst case, a company buys the tools, keeps the reporting, and blames the tools when nothing changes. Dropping a report is harder than buying a tool. Somebody's week was built around it, and somebody above asked for it.
Think big is what I say to others, and I try to do it myself. It is hard to think big with a head full of somebody else's spreadsheet.
The hours can come back, one report at a time. Whether they go to thinking, or to a longer report, is to be seen.