Skip to content

How I think about AI in organisations

I've never met an organisation short of capable people or good ideas.
And yet these alone aren't enough.

I believe an organisation should be clear about what it exists to do, what it wants to become and how AI could help it get there. That ambition needs to mean something for the people making decisions and changing everyday practice.

I conceived these models to give organisations language and structure for that thinking. They address three questions: what do we want to become, where are we on the adoption pathway, and what conditions are we creating for people to adopt technology and derive value from it?

What you want to become

Ambition Model

I developed the Ambition Model to help organisations be specific about what they want AI to make possible. It’s five questions that draw out a clear, organisation-specific reason for what AI should improve or enable. That specificity is what makes the hard choices possible, because every decision can be tested against what the transformation is actually for. It also gives your people a far better answer than “AI should make us more efficient and cut costs”.

Uniqueness

Does this make us more like everyone else?

Why this matters

When everyone can use the same tools, what sets your organisation apart? What do you understand that others don’t, and what can you do that they can’t? That difference is why people choose you. AI can amplify it, or quietly wear it away.

Choose one to see its question.

Five equal-sized petals: Uniqueness, Standing, Judgement, Effectiveness, Possibility. None is ranked. Choose one to read its question.

I believe AI is a tremendous opportunity. Efficiency, productivity and lower costs matter, but AI should also let an organisation do new things, do more of what it does well, and deliver on the reason it exists.

I use these questions in the Board-Ready AI Vision

That’s where you’re heading. Now, where are you really?

Where you are today

Adoption Pathway

Every organisation has its own reasons for using AI, and its own idea of what worthwhile progress would look like. I use the Adoption Pathway to examine how that progress develops: whether people have a reason to act, permission and support to change their practice, and evidence that those changes are helping.

The questions are shared. The answers, priorities and measures of success need to belong to your organisation.

Think about one team or part of your organisation. What has already been established? What still depends on a few enthusiastic people? Where does progress become difficult?

Those are useful places to begin. Progress should be judged against what you want AI to improve or make possible here.

Imperative

Give people a reason that matters here.

“We need to use more AI” leaves people to invent their own purpose. A clearer reason might be giving specialists more time to review difficult cases, or improving the service people receive.

What should become better here, and why does it matter now?

  1. ImperativeWhy does AI matter here, now?

    Give people a reason that matters here.

    “We need to use more AI” leaves people to invent their own purpose. A clearer reason might be giving specialists more time to review difficult cases, or improving the service people receive.

    What should become better here, and why does it matter now?

  2. AgencyWho has permission, confidence and clarity to act?

    Make it possible for people to act.

    Someone may be comfortable using AI for their own tasks and still have no permission to change a shared process. Leaders and managers need to make that permission clear and support people when they use it.

    What are people actually backed to change?

  3. ReadinessCan the organisation support real adoption?

    Give people the support to use AI.

    A team needs to know which tools and data it can use, how to check results, and where to take an uncertain question. Training helps when people can apply it with that support around them.

    Can people try something useful without having to resolve every uncertainty themselves?

  4. PracticeHow does AI become repeatable working behaviour?

    Make useful learning part of how the team operates.

    A useful experiment is worth sharing if colleagues can repeat it and know when to use it. Give them time to compare results and improve it together.

    What continues when the most enthusiastic person is away?

  5. InnovationWhat new capability and value can emerge?

    Use growing capability to try something new.

    As people become more capable, they may find opportunities beyond improving an existing task: a different service, a better decision process, or something the team previously couldn’t attempt.

    What could you now do that would be worth pursuing?

  6. OutcomesWhat changed, and was it worth it?

    Check what changed against what you wanted.

    Look at the quality of decisions and services, what people have learned, and whether any time saved went somewhere useful. The evidence should connect back to your reason for acting.

    What improved, for whom, and how do you know?

Where progress becomes difficult, I look more closely at the conditions people are working in.

What is standing in the way

Ten dimensions of AI adoption

I developed these dimensions to examine the organisational conditions behind adoption and value. An organisation can buy tools, run training and report growing usage while leaving those conditions largely unchanged. Capable people still need permission to change established workflows, and further training will not supply it. I believe meaningful, sustained results depend on addressing these dimensions together.

Transformative belief

Do leaders believe AI should change how the organisation operates, and are they prepared to act on that belief?

What you might hear

“We're already doing this.”

There's plenty of activity, with training, pilots and steering committees, but the way work actually gets done hasn't changed. Pilots succeed, then never scale, and nobody asks why.

I use these dimensions in the AI Adoption Reality Check

Contact

Start with the question you have.

You may have a direction to clarify, or adoption that has stalled. Get in touch about your organisation, and we can discuss whether I can help.