What Cameron Diaz and Jude Law can Teach us about AI Transformation

“You are, seriously, the most depressing girl I’ve ever met.”

I admit, not the most obvious place to start a blog about AI transformation, but stay with me.

In one of my favourite rom coms, The Holiday, Cameron Diaz’s character Amanda is trying to work out what happens next in her relationship with Graham, played by Jude Law. They live on opposite sides of the Atlantic, she is about to return to Los Angeles, and there is no obvious way of knowing how their relationship will work. So Amanda does what many of us do when faced with uncertainty: she tries to predict the entire future.

Her conclusion: they will fly backwards and forwards. Eventually it will become difficult. Work will get in the way. His children will struggle with it. They will argue. It will all fall apart. Everyone will be miserable. Graham’s assessment? “You are, seriously, the most depressing girl I’ve ever met.”

He’s not wrong, and, strangely, I think there is a lesson here for AI transformation, because right now we are doing an awful lot of trying to predict the ending.

Depending on who you speak to, AI is going to fundamentally reinvent how organisations operate, transform public services, automate huge parts of the workforce and unlock levels of productivity we have never seen before. The other view is that it is overhyped, unreliable, risky and destined to become another expensive technology programme that promised transformation and delivered a slightly better chatbot.

The slightly frustrating answer is that nobody really knows…

We don’t know what the ending looks like

I was reminded of this recently during a discussion I had with other consulting leaders working across the public sector. We talked about the extraordinary potential of AI, but also the difficulty organisations face in knowing quite how transformational it will ultimately be.

Will we genuinely redesign whole operating models around AI? Could we fundamentally change how services are delivered, how decisions are made and where people spend their time? Or will the reality be much more incremental?

Perhaps AI removes ten minutes from one process, automates part of another and helps somebody write a report a bit faster. Valuable, absolutely. Transformational? Maybe not quite in the way the keynote presentations suggested.

And there is another possibility: some of today’s excitement might simply turn out to be hype. After all, many people in today’s workforce lived through the dot-com bubble. Won’t AI be the same? The problem is that organisations are being asked to make decisions about AI today while simultaneously being expected to understand what AI will mean in three, five or ten years’ time. That is an impossible standard.

We don’t expect organisations to know exactly what their workforce will look like in 2035. We don’t expect them to predict every future customer need or political priority. Yet somehow with AI there is a temptation to think we need the whole destination mapped before we start travelling.

We don’t.

Transformation doesn’t require a crystal ball

Back to The Holiday. By the end of the film, Amanda and Graham still haven’t produced a five-year relationship strategy, an implementation roadmap or, disappointingly, a benefits realisation framework. What they eventually settle on is much simpler: “No set rules.”

They stop trying to work out every possible ending and start dealing with what is immediately in front of them. AI transformation needs a little more governance than a romantic comedy, admittedly, but the principle isn’t completely different.

Organisations don’t need certainty about where AI ends. They need enough clarity about what they should do next. That means putting the foundations in place now: being clear about the problems you actually want to solve, and then exploring how AI can enhance the solutions you devise to those problems.

It also means giving people permission and confidence to test things, run experiments, measure what happens, keep what works and stop what doesn’t. Most importantly, use what you learn to inform the next decision.

That is transformation.

It is not producing a perfect picture of the organisation of 2035 and then spending the next five years attempting to implement it unchanged. It is creating a clear enough direction, learning as you go and getting progressively better at making the next decision. You take every day as it comes.

Jump in. Just don’t jump in blind.

There are two equally dangerous responses to uncertainty: believe every prediction about AI and throw technology at everything, or wait until the future becomes clearer. Wait for the technology to mature. Wait for somebody else to prove the use cases. Wait until we understand what the workforce implications are. Wait until the future becomes clearer.

The problem with waiting is that the future only becomes clearer because organisations experiment with it. We build the capability to transform through doing: experimenting, learning, adapting and becoming progressively more ambitious about where AI can make a difference.

So, like Amanda and Graham, perhaps we need to get a little more comfortable with “no set rules”. That doesn’t mean jumping in blindly. It means having a clear direction, putting the right frameworks around us and accepting that we won’t know every step of the journey in advance.

Perhaps AI completely rewrites how our organisations operate. Perhaps it changes much less than today’s biggest predictions suggest. We’ll find out.

But we don’t need to know the ending before we start. Sometimes you just have to take transformation one day at a time.