Strategic Foresight vs Forecasting: They're Not the Same

"Can you forecast where AI is going in three years?" Half the time when someone asks me this, they want a forecast. Half the time, they want foresight. They're different things, and mixing them up has cost more than one strategy team a wasted quarter.

Forecasting

Forecasting tries to predict the most likely future. It assumes the underlying system is stable enough that the past is informative. It produces a single number, or a single curve with a confidence band. Examples: next quarter's revenue, next year's GPU demand, next decade's electricity load.

You measure forecasting by accuracy. The model that was right wins. The model that was wrong loses. Reality is a yardstick.

Foresight

Foresight tries to map the space of plausible futures. It assumes the underlying system might not be stable. It produces several scenarios, each with the conditions that would make it real. Examples: what happens if open-source models stay one year behind frontier models? What if energy policy makes large model training prohibitive in the EU?

You can't measure foresight by accuracy because the point isn't to be right. The point is to be prepared. Foresight wins when your organization has thought through a scenario before it arrives, even if a different scenario actually arrives.

Why the difference matters

Mistake one: asking for a forecast in a domain that's too unstable. You'll get a confident number that's wrong, and the team will plan against it as if it were right. Costly.

Mistake two: asking for foresight in a domain that's actually stable. You'll get four scenarios when you needed a number, and the team won't know what to do with them. Slow.

Which to use, when

Forecast when the past 5 years look like the past 10, and you have data. Demand planning, capacity planning, financial budgeting.

Foresee when the past 5 years don't look like the past 10, or you don't have data. Technology strategy, regulatory exposure, market entry into new geographies.

How they pair

The honest answer is: most real decisions need both. A capacity plan needs a forecast (most-likely demand) and a foresight overlay (what happens if a major customer churns). Strategy decks that only show one of the two are usually missing the more important half.

Forecasting tells you what's coming if nothing changes. Foresight tells you what to do when something does.

If you're hiring for either, ask candidates which one they think they're doing and why. The answer tells you what they'll actually deliver.