Start, get curious, get a little obsessed.

In 1609, Galileo Galilei heard a rumour. Someone in the Netherlands had built a device using two lenses that made distant objects appear closer. He had never seen it. There was no manual, no access, no invitation to try. He went and built one anyway. Then kept improving it until it could do something the original never could. Within months he observed four moons orbiting Jupiter. Nobody had ever seen them before.

He did not invent the telescope. He just refused to wait.

That gap between hearing about something and building with it used to take years. Knowledge moved slowly. Access was scarce. Working with a new technology required specialists, budgets, and time.

Not anymore. Building an app once meant designers, developers, months of back-and-forth. Today it takes hours. A working prototype, made by one person, in an afternoon. The work shifts from doing to directing, understanding what you want, and knowing how to get there.

Access is no longer the advantage. Everyone will have it. The difference will come from taste. From knowing what good looks like and what is actually worth building. Galileo did not replicate what he had heard about. He made it sharper, more precise, more useful. That is what is available to you now.

When Gutenberg's printing press spread across Europe in the 1450s, entire guilds of professional scribes faced the end of their trade. Some resisted. But many of the more skilled among them did something more interesting: they became editors, translators, and proofreaders for the new printing houses. Roles that had barely existed before. The skill had never really been copying. It had always been understanding the text. The press just made that distinction impossible to ignore.

We are in a similar moment. A reorganisation of what work actually is.

For a long time, your job was your identity. "What do you do?" is almost the second question people ask when they meet someone. So when that starts to shift, it feels personal. It is personal. It is also an opening.

The first things people outsource to AI are the tasks they already do. Email, summarising and brainstorming. That is a start. But once AI handles the less interesting work, the next question comes fast: does that work need a person at all? The first roles under pressure are the junior ones. Repeatable, low-risk, exactly the work AI is built for.

Go a few levels up and the picture gets more complicated. Middle management, focused on planning, monitoring, coordinating, drafting, that work is also largely automatable, once AI is woven into how an organisation actually operates.

But before anyone starts cutting those roles, pause. These are the people who know what is actually happening on the ground. They know the processes in detail, the people doing them, and where things break. That knowledge is exactly what you need to build an AI-native organisation. Do not remove them. Redirect them.

Work becomes less about your fixed position and more about what you can actually do. You will not necessarily work on one thing, or for one organisation. You will deploy yourself where it counts. The people who figure that out start doing more valuable work, closer to the reason they started in the first place. It is not about being technical. It is about judging what is good.

That asks a lot. The people who move well through this are curious and willing to be wrong in public. Not because those are nice qualities to have. Because the work demands it now.

Then there is the junior question. The traditional career ladder started with repetitive work. That is how you built judgment, by doing the boring stuff long enough to understand what good looked like. AI is taking exactly that layer. The tasks that used to fill the first two years of someone's career.

So yes, you will need fewer juniors in the traditional sense. But cut that layer too fast and you create three gaps that are hard to close later.

None of these gaps close themselves. They need deliberate structure. Rotate juniors through the work that still requires human judgment. Build AI literacy programs that run in both directions, not just top-down. Create space for juniors to teach, not just learn. The traditional apprenticeship model is broken. You cannot hire fewer juniors and assume the judgment will still be there in five years. It will not.

Look at your own organisation. Where are these gaps already forming? Which roles are being quietly hollowed out, and which ones are quietly becoming more important?

What do you actually want? Not what the job description says. What do you want to be known for?

That question is back on the table. Use it.

When Maya Angelou started writing I Know Why the Caged Bird Sings, she did not wait for perfect conditions. She created them. She rented a hotel room, no distractions, just a legal pad, a pen, and the decision to write. The story was not easy to tell. It meant facing trauma, silence, and experiences rarely given space at the time. She had an editor who believed it mattered. She did the work to make it undeniable. Resistance is not a warning sign. It is the weather serious work happens in. You do not get rid of it. You decide if you are going anyway.

Maya Angelou carried more than most people should. So it is not strange she felt resistance writing this book. But she was convinced it was important, that it could change something. So she did it. Not alone, but with help.

This is how we should approach AI. Different context, same dynamic. A lot of people fear losing their job, their identity. Insecurities come to the surface that nobody wants to show off. So it is easier to resist this new world than to enter it, even though it is there whether you want it or not.

When things change fast, resistance is the natural response. And right now things are changing faster than most people can track. Starting feels harder than waiting. But waiting has a shelf life. At some point the question stops being resistance versus learning. It becomes: how much ground are you willing to lose?

Start small. Make an app by vibe-coding. Try a tool you do not fully understand yet. It will feel awkward. Do it anyway. With everything available right now you can build things you never thought possible. You will get the hang of it. Then you will want more. And before long you are working differently, thinking differently, because of something you almost did not start.

You can be curious and afraid at the same time. That is fine. Do it scared. See what shows up.

Create a workspace and become part of our group. Every week we share one insight on what is moving in AI, so you stay close to it without having to follow everything yourself.

Before the compass, Polynesian navigators crossed thousands of miles of open ocean to find islands the size of a pinprick. They read stars, wind, wave patterns, and the behaviour of birds. They carried no maps. What they carried was an intimate knowledge of the ocean, and the judgment to read it as they went. The destination was always clear. The route was figured out along the way.

There is a lot of resistance to AI, but there is also pressure to do something with it. Maybe that is where you are right now. So you start with a tool someone recommended, or ask a colleague what they have been using. Before long you have tried five things. Some of it is faster, sure. But nothing connects. It starts to feel like AI is not delivering what you heard it would.

But you probably already know that is not quite true. Picking up tools at random and hoping things fall into place is not a strategy. Like the navigators, you need a destination before the route makes any sense. Be open to changing course. But know where you are going first.

What does your ideal department or organisation actually look like? How does it work? Set your North Star, the end goal that makes every smaller decision easier to take.

Once that is clear, you can start building toward it. With an open mind, because what is possible today can be outdated in weeks, and what feels impossible now might be straightforward in a year. That is not a reason to wait. It is a reason to pick a direction and keep reading the signals.

So what is your star?

Henry Ford did not invent the moving assembly line in isolation. He studied the meat-packing industry in Chicago, where disassembly lines moved carcasses past stationary workers. He inverted the idea. What looked like a radical transformation of manufacturing was actually a recombination of existing logic, applied with total commitment to a new context. The Model T was not the innovation. The way the organisation worked was the innovation.

The work remains. How we work changes.

Which 15 skills should you launch the marketplace with?

Before Michelangelo painted a single figure on the Sistine Chapel ceiling, the plaster failed. Early sections developed mould, had to be stripped back, and redone. He spent weeks adjusting the mixture and getting the surface right. He wrote about it at the time. He was furious. The ceiling has been there for five hundred years. That is what the preparation bought him. The unglamorous work is almost always the work that decides everything. Not the painting. The surface it sits on.

Before you reach for the next shiny tool, define the project. Map the workflows in detail. See where tasks overlap, where handoffs happen, where things currently break. Know what needs to happen before you decide what to build it with.

Once you understand the full process, the tool question gets much simpler. A combination of Notion, Slack, Claude, and Make might be enough or a good start. Building something from scratch might be faster than buying an off-the-shelf solution. The top-of-the-bill app might actually be right.

All are valid. But the answer only becomes clear after the mapping is done. Without it, the shiny object trap closes around you. You pick something that does not fit, spend weeks on workarounds, and switch tools before anything has had time to work. The problem was never the tool. It was that the surface was not prepared.

What you need is a system that fits how your team actually works and holds up during a busy week. Simple enough to keep running when things get hard. Stay curious about new tools. Good ideas arrive that way. But know what your project needs before you go looking. That knowledge is your filter. That is your plaster.

During the Crimean War, Florence Nightingale kept meticulous records on every soldier who died in the British military hospitals. The numbers revealed something the army had not expected: far more men were dying from preventable infections than from battle wounds. She turned those findings into a polar area diagram, one of the first data visualisations used to drive public policy, and presented it directly to Parliament and the War Office. Sanitation reforms followed. Mortality rates collapsed. The data did not speak for itself. She made it impossible to look away from.

The data is already in your organisation. Financial records, CRM, HR, marketing, sales, operations. It exists. The question is whether it gets used, or whether decisions still run on gut feeling and whoever happens to have the right spreadsheet.

Gut feeling is not always wrong. But it has a ceiling. AI can do what Nightingale did, continuously and across every system at once. Surface the patterns. Present what would otherwise stay buried. Flag what is likely to happen next, so you are anticipating rather than reacting.

But even with all the data connected, and honestly in most organisations it is not, you still need to ask the right questions. That part does not get automated. Without a clear hypothesis, AI produces noise. The thinking and direction has to come from you.

When good data is actually used, internal politics quieten down. The argument shifts from opinion to evidence, and decisions move faster because of it. Start with what you have. What data exists in your organisation right now, and what is the first question you would ask if all of it was available to you? That question is worth writing down.

Between 1846 and 1854, Charles Darwin studied barnacles. Eight years. Friends urged him to publish the evolution theory he had already developed. He ignored them. He wanted to understand one form of life in full before saying anything about all of it. When On the Origin of Species appeared in 1859, it held. He had read the fingerprint before he drew the map.

Most organisations go straight to the tools. They start using AI without first looking at themselves. What kind of organisation is this. What drives the culture. Where the resistance lives, and where the energy already is. The answers to those questions determine whether anything you build will actually land.

Over time, every organisation's culture shifts. Sometimes it moves toward something better. Sometimes it drifts quietly away from what made it work. In a moment of significant change, that gap matters more than most leaders want to admit.

AI adoption is a culture problem as much as anything else. How curious is the organisation? How does it handle the first failure? These are measurable qualities, and they tell you exactly where to start.

The organisations that adapt well share one thing: they know what they are before they try to change. They understand where energy flows. Without that knowledge, tools get abandoned after the first friction. And with AI, there will be friction.

What do you actually know about your organisation's culture right now? What are you assuming without evidence?

That is your fingerprint. Read it before you draw the map.

Before her 1932 solo Atlantic crossing, Amelia Earhart spent weeks on the ground. She mapped fuel load, weather windows, aircraft mechanics, route options. She knew what she could control and what she could not. When the conditions looked good enough, she went. Fourteen hours and fifty-six minutes later she landed in a field in Northern Ireland.

The flight was the easy part. The assessment made it possible.

There is no perfect moment. There is just an honest read of where you are: what your culture can absorb, what your data supports, what you are actually trying to do. You do not wait until every risk is gone. You look clearly at what you can control, accept what you cannot, and go.

An AI opportunity matrix makes that call concrete. It maps your projects across two axes, impact on the organization, and readiness to execute, and tells you where to start, what to protect, and what to leave alone. Impact scores on strategic relevance, business value, and learning potential. Readiness scores on feasibility, repeatability, and manageability. High on both means top right. That is where you go first.

So map your ideas, workflows and tasks to make it visible. Where do you start?