AI keeps winning you the quarter. Is any of it building the decade?

Look at the AI business cases stacked on your desk. Almost all of them say the same thing in different words: this will cost less, or this will be faster. Every one is a genuine win, and every one is about now. And somewhere beneath the approvals a slower question waits, the kind that does not fit on a business-case template: is any of this actually building the company we want to be in ten years, or are we simply making this quarter look a little better, again?

Here is a reframe worth sitting with. A 2025 systematic review in the Journal of Innovation & Knowledge, drawn from 1,377 studies, argues that AI's most valuable contribution is not efficiency at all. It is the capacity to help organisations escape short-termism: to allocate resources better, to engage stakeholders more genuinely, and to build toward long-term growth, even toward the Sustainable Development Goals. The efficiency wins are real. They are also where most organisations stop, and stopping there is the trap. The same technology that shaves a cost this quarter can, pointed differently, build an advantage that compounds for a decade. Which one you get is not a property of the AI. It is a choice about what you aim it at.

Why is short-termism the real risk, not the tool?

Because the evidence shows AI paying out mostly in the near term, and near-term wins are seductive precisely because they are so easy to book. A second large review, of 2,670 studies in Frontiers in Artificial Intelligence, found AI's measured value skews toward short-term operational improvements, while sustained strategic effects remain scarce and under-proven. Read the two reviews together and they say something uncomfortable: we have become very good at using AI to win the quarter, and we have little evidence yet that we are using it to build the decade.

The gap is not capability; it is intent. It is the same pattern behind why most organisations fail at AI adoption, seen from a different angle: the tool works, and the human system points it at the nearest, safest, most countable win rather than the one that matters most.

What does it look like to point AI at the long term?

The Journal of Innovation & Knowledge review is specific about the roots AI can address, and each one translates into a different way of aiming it.

• Resource allocation. Use AI to see where capital, attention and effort are genuinely best spent over years, rather than to squeeze the current cost line a little harder.
• Stakeholder engagement. Use it to understand and serve customers, employees and community more deeply, not to automate them into silence.
• The horizon itself. Use it to model and build for futures, not only to optimise the present.

The mechanism the review names for turning this into lasting value is knowledge creation and dynamic capabilities, the discipline of sensing what matters, seizing it, and reconfiguring the organisation around it. That is the opposite of a one-off saving. It is stewardship: managing the business as something you intend to hand on in better shape, not only something you report on each quarter. It is also the deeper reason you cannot out-hustle the machine, because hustle optimises the present, and stewardship builds the future.

How do I build the long game into how we use AI?

By making the long horizon a deliberate part of every AI decision, rather than a virtue you hope emerges. A few habits do most of the work.

1. Put a decade question on every AI business case. Beside "what does this save this quarter" write "what does this build over ten years". When the second box is empty, notice it.
2. Aim some AI at allocation, not just efficiency. Use it to decide where to invest over years, which is a harder and far more valuable question than where to cut.
3. Use AI to hear stakeholders, not mute them. A deeper understanding of customers and employees compounds; automating them out of the conversation quietly hollows the business.
4. Convert wins into capability. Treat each efficiency gain as raw material for a durable capability through the sense, seize, reconfigure loop, or watch it fade.
5. Govern for trust as a long-term asset. Accountability, transparency and privacy are what let an advantage last, and the review names them as preconditions rather than niceties.

What does this ask of me as a leader?

A shift of horizon, and of care. Efficiency is the easy, measurable, near part, and it will always shout loudest in the business case. Stewardship is the harder, slower, more human part, and it is where the durable value actually lives. The move the evidence quietly asks for is to stop managing AI to make the quarter look good, and to start using it to build something worth inheriting.

That is not a softer ambition than efficiency. It is a larger one. Point AI at the decade with the same seriousness you already bring to the quarter, and you will find you are building an advantage of the kind that outlasts the old way of working, still compounding long after this quarter's saving has been spent and forgotten.

About the author

Thomas Green

British technology futurist, AI keynote speaker and advisor. Thirty years across enterprise technology and AI strategy, helping leaders navigate the future of work. The futurist who died.