Skip to content
AIMAI

Our thinking

More AI doesn’t automatically make a smarter business

AI can make individuals more capable while leaving the business more fragmented. The leadership challenge is to bring that intelligence together.

Jonathan BradleyPublished

Give everyone in a business access to AI and you would expect the business to become more intelligent. It sounds fairly obvious.

I think there is another possibility worth paying attention to. Everyone becomes more capable, but the organisation becomes harder to hold together.

Better answers, separate conversations

Someone in sales has an AI helping them prepare for a client meeting. Operations has another helping with schedules. Finance is exploring the numbers. Each person has useful information, a helpful conversation and a clearer idea of what to do next.

But those conversations may never meet.

The meeting assistant knows what was discussed. The email assistant knows the inbox. The reporting tool knows the figures. What one person has learned does not automatically become available to the next.

That is what I mean by the Many Caves Problem. There can be a great deal of intelligence in each cave, with very little passing between them.

The person in the middle

Usually, somebody has to make the connections. They remember the conversation, copy the information, explain the background again and point out that a decision has already been made.

That work can be easy to overlook. We notice the time saved producing an answer. We are less likely to notice the effort involved in getting everyone to work from the same understanding.

There is another complication. An AI conversation often contains more than the final answer. It includes the questions, the assumptions, the options that were rejected and the reasons behind a decision.

If only the answer is shared, much of that context stays with the individual. When somebody changes role or leaves, the business can lose access to the thinking that helped shape its work.

More capability needs more direction

None of this is an argument for holding people back. The opportunity is too useful for that.

It does change the leadership question. Alongside asking what AI can help each person do, we need to ask how those individual gains will help the business work better together.

What are we trying to achieve? Which information should we trust? What needs to be shared? Who decides when two perfectly reasonable recommendations pull in different directions?

More confident people do not automatically produce a more aligned organisation. They may simply become better at pursuing different priorities.

Leadership still has to provide direction, make room for disagreement and bring the right people into decisions. Access to AI does not do that work for us.

What should the business keep?

Before adding another tool, I think it is worth asking what will remain with the business after the task is finished.

Will the useful knowledge be available to the next person who needs it? Can someone check where an answer came from? Will a new colleague understand the decision, or just inherit its outcome?

That does not mean every conversation should be visible to everyone. Some work needs to stay private. Sharing has to respect the people involved, the purpose of the information and who should have access.

It does mean making those choices deliberately, rather than discovering later that important knowledge has been spread across personal accounts and separate tools.

There is a technology job here: connecting information and making it available in the right context. There is also a leadership job: deciding what matters, who needs to know and how the organisation learns from its work.

I find the second question at least as interesting as the first.

AI is giving more people access to more intelligence. Whether that makes the business wiser depends on what we do with it together.