Organisations Are Getting Smarter.

That Doesn’t Mean They’re Thinking Better.

Over the past year, there has been a noticeable shift in how organisations talk about AI. The conversation has moved on from tools and experimentation, from pilots and possibilities, towards something a little more grounded. Revenue, efficiency, decision speed. The kinds of outcomes that show up in performance rather than presentations. It feels like progress because the language is more commercial, more serious, and more closely aligned with how businesses actually operate.

But it still isn’t quite on the money.

A growing number of studies and articles are pointing towards a similar pattern. Research from organisations like Boston Consulting Group highlights the gap between companies investing heavily in AI and those actually realising sustained returns. At the same time, writing in Harvard Business Review continues to return to familiar themes, execution challenges, alignment issues, and decisions that don’t translate cleanly into action. The language varies, but the direction is consistent.

Organisations are becoming more intelligent, but not necessarily more effective.

The accumulation of intelligence

What is interesting is not that this gap exists. Organisations have always struggled to convert insight into action. What feels different now is the scale and speed at which it is appearing.

Intelligence is no longer scarce. Data flows continuously, often without friction. Systems surface patterns before they are visible to teams. Models generate predictions that would have taken weeks or months to produce even a few years ago. In theory, this should lead to better decisions, faster responses, and stronger outcomes. It should sharpen how organisations operate.

In practice, it often creates something else entirely.

More signals appear across the organisation. More interpretations are offered. More possibilities are generated. But instead of clarity, there is often a subtle increase in hesitation. Decisions take longer, not because there is a lack of information, but because there is too much of it, arriving faster than the organisation can absorb or interpret. Intelligence accumulates, but momentum doesn’t always follow.

The language above the problem

Most of the language used to describe this still sits one level above what is actually happening. We talk about execution, about transformation, about adoption and alignment. These are all valid descriptions, and they capture something real. But they tend to describe the symptoms rather than the mechanism underneath.

• Execution breaks down because decisions are unclear.
• Transformation stalls because ownership is diffused.
• Alignment becomes difficult because interpretation varies across the system.

Each of these points to the same underlying issue but rarely names it directly.

How organisations actually think

What these signals are circling around is something more structural, something that sits beneath process, beneath strategy, beneath even culture. The ability of an organisation to interpret, prioritise and act on intelligence.

In simple terms, how the organisation thinks.

Not in an abstract or philosophical sense, but in a very practical one. Who sees the signal first. Who decides what it means. Where decisions are made, and where they are deferred. How quickly action follows, and how consistently it aligns with intent. These are not small details. They are the mechanics through which value is either created or lost.

For a long time, organisations did not need to think particularly well. They needed to execute consistently. Processes were designed to reduce variability. Systems were built to standardise decisions. Information moved slowly enough that centralised control remained effective. Judgement could sit at the top, and the rest of the organisation could operate within defined boundaries.

That model worked, because the environment allowed it to.

Where the model starts to strain

The strain begins when intelligence starts to move faster than the structures designed to contain it. AI increases the volume and velocity of information inside the organisation, but it does not automatically change how that information is interpreted or owned. The underlying decision structures often remain intact, even as the environment around them accelerates.

This creates a subtle but important tension.

Signals begin to appear everywhere, across teams, functions and systems. But the authority to act on those signals does not always move with them. Decisions continue to escalate, even when the information needed to make them is already present. Interpretation becomes fragmented, as different parts of the organisation see different versions of the same reality. The organisation becomes more informed, but not necessarily more decisive, and this is where the friction shows up.

Insight accumulates. Action slows.
Value leaks and is often unnoticed.

From intelligence to cognition

What is emerging is not just a technology gap, or even a strategy gap. It is something more fundamental.

A cognition gap.

The difference between having access to intelligence and being able to think with it as an organisation. The difference between generating insight and allowing that insight to shape behaviour in a coherent, timely way. This is not about more data, or better models, or faster systems. It is about whether the organisation itself has adapted to operate in an environment where intelligence is abundant.

I explored this more directly in a recent piece, When Organisations Learn, which looks at how organisations move beyond processing information to actually adapting based on it. Because learning, in this context, is not about accumulation. It is about change. About whether intelligence alters decisions, and whether those decisions translate into action.

As AI continues to expand the amount of intelligence available, the constraint is shifting.

It is no longer access to insight. It is the organisational ability to think with it.

And that is a different challenge altogether.

For some organisations, that question stays theoretical for a while.

For others, it becomes more immediate.

Where does intelligence currently enter the organisation?
Where does it stall?
And who is actually able to act on it?

If that’s a question worth exploring, the Decision Architecture Review is a practical place to start