Ten years ago the internet ate my lunch. Now AI is coming for dessert.

In 2016, before LinkedIn was a corporate amphitheatre and AI became everyone's strategy, I wrote a blog post saying, "Digital is 10% tech and 90% human. But organisations talk about it as if it's 90% tech and 10% human."

The hypothesis appears to have aged well - here’s why.

This was my view from inside The Times, where I had spent most of the previous decade leading the move from print into digital. The line seemed to resonate with what many were feeling but hadn't quite named. When I published the post on Medium (basically, ye olden days Substack), people I hadn't spoken to in years got in touch to ask if they could use it in their board presentation.

It boomeranged back to me and I heard it quoted in rooms I had never been in.

Ten years on, and the same hypothesis can be applied to a different technology shift.

Sooner or later, every company has to hold the meeting it has been postponing. The one where someone says the shadow fear out loud. If we were starting this today, would we build it like this?

When the editor called a small group of us to the bunker room at The Times back in 2009, we all had the same thought. Holy shit.

Two centuries of print headed for the skids. A free website about to start charging in a move so counter-cultural it was on the news. A bunch of us, in this room, tasked with working out what came next.

The same meeting is being held in scale-up boardrooms now. Politer vocab (hopefully). AI as the disruptor du jour. Same question underneath.

If we were building our brand from a blank page today, what would it be?

And the follow-up that sticks in the throat. How much of what we have already will survive in the new world order?

I have worked with teams grappling with this question for over twenty years. What I have come to understand is that there are three layers beneath it. Yet the language we use usually focuses on one. 1/ The strategic layer is the one boards discuss - are we focusing on the right thing? 2/ The operational layer is the one teams feel - is there enough momentum to make a shift? 3/ The third layer is the one that is felt and influences everything we do - are we resourced (in every sense) to make this shift?

The questions organisations are now facing sits at the same crossroads.

1/ Is the question we're answering still the right one?

By 2014, a chart made by media commentator Kevin Anderson was being passed around newsrooms across Britain and America. US newspaper advertising revenue from 1950 to 2014. A long, broadly upward climb to about $67 billion in 2000, then a near-vertical drop. Underneath, two lines climbing the other way. Google and Facebook ad revenue.

The chart wasn't really about newspapers at all. It was about what happens when the question your business was built to answer stops being the question the market wants an answer to.The business model - turning eyeballs into advertising revenue - was disappearing in front of our very… well, eyes. The money that had funded journalism was moving to two American technology companies (who didn't pay journalists). And worse, the journalism we produced was helping fuel their profits.

We were solving the wrong problem with extraordinary skill.

In 2017, I did a talk quoting a ‘break glass’ statistic: A major global financial news publisher was already producing 65% of its articles through automation. That number is now nine years old. It was the early sound of a much bigger shift, and most boardroom conversations at the time were still about content marketing strategy.

There is a useful name for this. Some people call it epistemic debt. The accumulated cost of making decisions inside a frame that no longer matches reality. I think of it more plainly. It is the cost of running a business that is operationally answering the wrong question, with excellent execution, using an outdated map.

It can be hard to spot until it's too late (that famous Nokia tale favoured on MBA programmes). It starts to show up as misalignment. Or teams pulling hard in different directions. AI initiatives multiplying without any clear aim. A leadership team looking confident in front of the board while privately knowing it's a punt. Metrics looking good, while the existential creep follows you around like a bad smell.

The first symptom of the mismatch is almost always a sense that the organisation is working harder for less internal traction. By the time the strategy itself visibly fails, the cost is already eroding the bottom line.

The 2026 AI version of the ‘internet ate my lunch’ chart isn't being printed and passed around (or if it is, I've missed it, so please get in touch).

It is certainly showing up in cap tables. In mass staff culls. In off-the-record investor conversations about whether the moat the business was funded on still exists. In the thought experiment some boards are now brave enough to run. Would we still build what we have today?

Most businesses funded between 2018 and 2023 were backed on a market thesis written before generative AI had hit its stride. Some still answer the question they were built to answer. Some no longer do. Others do it less well than their new competitors, and at much higher cost.

The Times caught on early in that first epoch of digital disruption. It was a brave move, and one I have taken huge learning from on a human level. Because it's the humans who decide whether they can engineer enough honesty in the room to ask the really difficult questions (without it costing someone their job).

2/ Can we get there from here?

It starts with what I call ambient malaise. Something you've been chewing over for months. Not in strategy off-sites. More like on the M1 at too-late-o’clock. The business needs to reinvent itself faster than the operating model can manage, and probably differently from what your investors have been hearing.

Now you have to say it out loud, and get backing.

This is where the next question arrives, and it almost always blindsides the people in charge. The business is optimised for a reality that's disappearing. The operating cadence, the metrics, the team structure, the reporting rhythm. Now that same set up is trying to run two different programmes at once. Keep delivering on targets. Invent the business of the future.

This was the underbelly of the harder conversations I was having over and over at The Times across the years that followed the paywall decision. The new business model was bold. But the hard work was getting it to land. We had to become a different kind of business while continuing to publish a newspaper every day. The journalists who had written for infinite web audiences were now writing for infinitesimal numbers of subscribers. The advertising team that had sold reach was now selling relationships in a market not built for that framing. The metric of choice changed from drive-by viewers to deep engagement. The default unit of work changed from waterfall projects to small iterative teams.

It felt like a Sisyphean task.

The cost of that renegotiation was high.

A friend asked me at the time what I did. I said:

Basically get journalists to do stuff they don't really want to do, while explaining that their reader base had evaporated.

There was anger, confusion, skepticism, anxiety. The full gamut. Questions of meaning, status and identity. All of it adding up to organisational friction that everyone felt and absorbed, but didn't show up on a spreadsheet.

Sure, it also eventually showed up as a healthier P&L and pride in what the team had built. But the path from the before times was treacherous, winding and very very long. From the inside it felt like cutting through fog with a spoon.

Most strategy decks underestimate this. The CEO can change their mind in a week. The deck can change by next quarter.

The organisation cannot. It needs six to eighteen months to reorient around a different answer to its founding question. During those months, productivity dips, attrition rises, and the leadership team holds coherence in public while privately holding uncertainty. All the while, new market entrants nipping at their heels and customers in mid transformation themselves.

Most leadership teams I work with a sense of what they need to do. The work isn't the ideas. It's helping them set the direction together, and board the same bus.

3/ Am I still the right one to lead it?

When I came back from my first maternity leave in 2009, I walked into a job that wasn't quite the job I'd left. Same title. Different work. The role was rewriting itself in real time and the official version hadn't caught up.

The cost of that gap was the bit that followed me home. It didn't stop when I shut the laptop.

The tangible stuff - strategy, operations, tech - gets talked about. The third question doesn't, but it is certainly felt and it comes at a personal cost.

The business you started is not, quite, the business you are running now. The expertise you built your career on is not, quite, the expertise the organisation now needs. The role you were hired for is no longer the role you're actually doing.

How the cost shows up is mostly invisible. It looks like coping. It looks like leadership.

This is what 10% tech, 90% human speaks to. AI may be accelerating the thinking. But the ones with the pulse - us - we’re still carrying the weight of the transition. And it is not a soft cost. It is the speed-limiter on every other decision and deliverable in the building.

It is also the question organisations are least good at hearing. Strategy gets noticed. Operational friction gets noticed.

The cost being absorbed by people almost never does - until it shows up as a slow turning-away that no one names but everyone eventually feels.

Most organisations do not fail because they miss the signal. They fail because the system cannot metabolise what the signal means quickly enough.

That is why the AI challenge was never mainly technological. It is, and always has been, profoundly human.

_____________________

Over to you

This piece names three questions most leadership teams are circling and most articles don't get to. Reading them is one thing. Working with them is another.

I have been grappling with how to bring it all together, and it’s resulted in me developing a tool - The Margin Read - that helps leadership teams read these three questions in their own organisation, with their own data, in their own language. It surfaces what's going on that no one has named. It puts the unspoken into a form a board can work with.

I am running a selective cohort of CEOs, boards, exec teams and founders through the tool over the next three months. The cohort gets the diagnostic, the read, and a working session with me on what surfaces. In return I get sharper feedback from leaders living inside this terrain right now.

If you have recognised your own organisation in any of the three questions above, that's the qualifying signal.

Email me at lucia@theboldmargin.com with a line about what's live for you. No sales pitch - just an honest exploration.