Foresight as an instrument, not a report.

Somewhere in your strategy is a sentence that begins this way. Most teams have never uttered it aloud.

The future that matters to you is the handful of ways the world could turn that would make your plans wrong. Find those, and you know exactly where to focus your attention. Skip them, and you can read every trend report ever written and still be surprised.

So foresight begins with your team, not with the world. We finish that sentence as many times as it takes, sort what is settled from what is still a gamble, and write each gamble down with what evidence would settle it. That short list is the instrument. Everything after is keeping it honest.

There is more published about the forces shaping the world every week, in more languages, than any analyst desk could get through, and most of it isn’t news. It is the statistics, filings and prices that institutions put out because they must. A machine can read all of that, in the language it was published in, keep a copy as it was that day, and compare what it found with what we agreed would count as relevant movement. It is cheap work for a machine, and it is checkable: every line it produces arrives with the source it came from.

What a machine is not for is deciding what any of it means. It drafts. My team and I read, change and release under my name — or we don’t. Which way the evidence leans, how thin it is, and what you might do about it are the judgments a person is paid for, and they stay with one.

It is now easy to make a machine produce a view of the future. Ask it for four scenarios and you will get four. Ask it for a probability and it will print one. Ask a dozen copies of the same model to look through a dozen different lenses and they will disagree, politely, blinded by the same blind spots, then agree on a story at the end.

Generating a picture of the future has become cheap. Checking one is as expensive as it ever was, and the checking is the whole job. So I use the machine where its work can be checked (reading, copying, comparing, drafting) and keep it out of the places where it can’t (deciding what matters, deciding what it means, putting a number on what nobody can know). And what counts as movement is agreed before the data arrive. That is what lets a quiet month be a finding, and what stops any of us, me included, from reading into the evidence what we want to believe.

The first season is a few big uncertainties, one team and a short watchlist.

It is built to grow: more markets, each with its own gambles; sources beyond the official ones, the filings and the cultural signs that move before the numbers do; your own people sending in what they see, with the source attached. And a record that accrues. Every gamble is written so that it can resolve, and when one does we will see which signs led, which misled, and which of your team’s own assumptions the world confirmed.

A team that keeps this up for 12 to 24 months will start to see something almost no other team can: which of its instincts about the future to trust.

None of that record exists yet. This work is set up from the first day so that it can.

Navigating is what you do when the map in your hand still gets you where you want to go. If that’s where your team is, I’ll say so in our first conversation and we’ll stop there.

This work is for the other case: a decision that matters, resting on a future you can’t see, and a team that suspects its plans assume more than anyone has said aloud, and wants to hold that up against what’s arriving.