You Can't Learn to Swim by Reading About Water

A new series about transforming from AI aware to AI ready.

I know. It's the kind of headline that peppers every other LinkedIn post and makes your toes curl. Bear with me though as I can't think of a better way to frame the AI problem most of us are wrestling with. And it's a big problem because it's not hyperbole to say that our jobs depend on it.

Most of us have watched the keynotes, seen the demos, tried the tools. We can probably talk a good AI game. However, if someone asked you tomorrow to commission an AI project, assess a vendor or redesign a workflow around it, you would undoubtedly feel a sense of rising panic.

the issue is that if you haven't built anything with AI, you haven't felt the speed of it or made the mistakes that teach you where the boundaries are. You haven't had the moment where it produces something unexpectedly useful and you think right, now I get what this is for.

That's not a knowledge gap, it's an experience gap. And at some point you have to stop reading and start using, so that you understand the art of the possible. You need to get in the pool.

AI aware is not the same as AI ready

A good client and friend of ours works for a large player in financial services. Thousands of people, massive profits. At an event in December he pulled us aside and said, quietly, like he was admitting to something, "None of us really get AI, we're just bluffing every day".

He's not alone. We hear some version of this confession every week from all sorts of people in all sorts of industries. Really smart and capable people managing big teams, businesses and budgets. Every one of them is AI aware. Almost none of them feel AI ready.

The reason isn't that they're behind what's being reported. The reason is that most of what you can read about AI starts with the technology. Here's a large language model. Here's how tokens work. Here's a prompt engineering framework with seven steps and an acronym. It's like teaching someone to swim by explaining fluid dynamics.

What you actually need isn't AI expertise. It's AI fluency and the difference really matters. Expertise means you can build the thing. Fluency means you know what the thing can do, when to reach for it and how to talk about it with your colleagues. You don't need to understand how a large language model works any more than you need to understand the chemistry of chlorine to get your laps in.

We've been here before

Just over ten years ago, we found ourselves at Freeformers, a startup revealing what digital skills could do for leadership teams at the likes of News UK, the BBC, Tesco and Barclays. We taught CEOs to code a working app in a day. We were part of the genesis of the Barclays Digital Eagles programme. The whole point was the same - don't explain the technology, let people experience what it can do. The confidence naturally follows.

Today's shift feels similar but more existential. Back then, digital transformation was about getting ahead. AI feels like it's as much about survival. The pace is faster. The capability gap is wider. And the cost of watching from the sidelines is higher than it was a decade ago.

Two frameworks for every AI decision you'll face

How then, can we become fluent and confident in a way that isn't all about the tech?

Well, our approach us expressed in two frameworks that give you a usable and stable mental model for every AI use case you'll encounter. One for generative AI and another for agentic AI. The tools will keep changing, but the categories within them won't.

Generative AI is the bit everyone's tried. You type a prompt, you get a response. You're in the driving seat the whole time. Powerful, but manual. Like having a brilliant assistant who only works when you're standing over them.

Agentic AI is what's coming next. You set the goal, define the boundaries and decide when it should escalate back to a human. Then it runs. Whether you're there or not.

These frameworks aren't designed for the people with technology jobs. They're designed for the people in the business who need to understand AI well enough to commission it, evaluate it or lead it without having to become a technologist.

Over the coming weeks we'll dive into the detail. For now, here's how the two frameworks map the landscape.

The Stuff Framework (Generative AI)

Inspired by this LinkedIn post last year which was itself a build on this OpenAI paper this is abut classifying the tools you prompt to produce things. You ask, it creates. You direct, it delivers. We call them "Stuff" because the language should be as accessible as the technology is becoming.

• Create Stuff: from blank page to finished output

• Find Stuff: surface what matters, fast

• Build Stuff: make functional things without a dev team

• Make Sense of Stuff: turn data into decisions

• Think Stuff Through: a thinking partner on demand

• Do Stuff Automatically: set it up once, let it run

The Keep Framework (Agentic AI)

The systems that operate on your behalf, continuously, without you standing over them. You brief an agent the way you'd brief a capable colleague. Then you walk away and it keeps going.

• Keep Watch: eyes on everything, all the time

• Keep Order: everything in the right pile

• Keep Moving: the work flows, even when you stop

• Keep Talking: always available, always on-brand

• Keep Connected: everything talks to everything

• Keep Learning: smarter tomorrow than today

The language is deliberately informal. When frameworks use plain language, people remember them. They use them in meetings. They apply them without needing a reference guide. The moment you have to look something up, the framework has failed.

Most teams will start with Stuff (generating outputs) and graduate to Keep (delegating operations) as their confidence grows. The two frameworks work together. They give any team, in any industry, a shared vocabulary for the whole landscape of what AI can do.

What's coming in this series

Over the coming weeks, we're going to explore both frameworks in detail. Not as theory. As practical guidance you can take into your next team meeting.

Next up: a deep dive into the Stuff Framework. What each of the six categories means in practice, where teams typically start, where the biggest quick wins are and the mistakes that waste the most time.

Then: the Keep Framework. How agentic AI changes what's possible, what it means to delegate to an AI agent and how to know which of the six Keeps your organisation should prioritise first.

After that: twelve posts, one for each of the twelve components. Real examples. Real use cases. The kind of detail that turns understanding into action.

If you've been feeling like AI is moving fast and you're not sure where you fit in, this series is for you. Not because you need to become technical. Because you're already capable of leading with this stuff. You just need the right map.

Subscribe so you don't miss the next one. And if you know someone who's been quietly feeling behind on AI, send this their way. They're not behind. They just haven't found the right starting point yet.

It's time to get in the pool.