10. I Cannot Put My Name On Something I Do Not Believe

I have been thinking about the Lebanese man.

Not a specific person.

A composite.

The businessman who arrived on shores that were not his, built something, failed at parts of it, kept going, passed it to his children, who passed it to theirs.

What I keep trying to locate is the source of that fearlessness.

Because it is not recklessness.

He is not careless.

He just does not seem to need permission from anywhere outside himself before he moves.

I have spent a lot of my life waiting for permission.

The MBA felt like permission. Paying for it daily for ten years felt like the price of permission. Getting to the other side of it and still not quite being able to step fully into what it represented felt like discovering that the permission I had been buying was not actually for sale anyway.

I think permission was never going to arrive from the outside. I think I knew that and kept looking there anyway because the alternative, simply deciding you are ready, simply putting your name on something and standing behind it, is one of the most exposed things a person can do.

Anything my name is attached to has to be excellent.

I did not choose that standard.

It chose me.

And it is as much a burden as it is a value.

I do not say things I cannot take back.

I do not make moves I cannot defend in a year, five years, ten years.

I do not publish what I am not sure of.

And so before anything moves I engineer it and re-engineer it and think it and rethink it until either it is ready or the moment has passed.

This is the perfectionism trap. I know it well. I live in it.

And it is the context for everything I am about to say.

I work in AI.

I help organisations and the people inside them understand what AI actually is, what it can realistically do, and what it means to adopt it responsibly rather than just enthusiastically.

I have been doing this in rooms, on stages, on screens, with financial services professionals and corporate teams and individuals trying to figure out what this thing means for their working lives.

I do this work seriously. I care about it deeply. And I carry a set of fears about it that I have not said out loud publicly.

Until now.

The first fear is about tools.

There are too many of them. More arriving every week, each one claiming to be the thing that will change how you work, each one requiring its own learning curve and its own subscription and its own mental real estate.

I worry constantly that the people I work with will walk away from a session genuinely motivated, try three tools in the first week, find the experience fragmented and exhausting, and quietly return to doing things the way they always did. Not because they are resistant. Because nobody designed a sustainable path for them to follow.

I wrote about this elsewhere. I will keep writing about it because I have not yet found a satisfying answer.

The second fear is harder to say.

When I walk into a corporate organisation and recommend a tool or a process or a way of thinking about AI, I am recommending it based on my best current understanding.

I am not a technologist.

I am someone who has studied this carefully, tested things extensively, and developed a set of instincts I trust.

But instincts are not guarantees.

What happens if something I recommend goes wrong? What if the tool hallucinates at a critical moment? What if the process I helped design creates a dependency that costs someone their job or their data or their client’s trust? What if I was wrong and an organisation built on that wrongness?

I raise this not to be dramatic but because I think;

The organisations I work with are not running experiments. They are running businesses, going concerns; with real people and real stakes. That deserves more than enthusiasm. It deserves caution.

So I ask questions that slow things down. I say we need to be sure rather than think. I check the data infrastructure before I celebrate the interface. Some people find this refreshing. Occasionally people find it inconvenient.

I find it necessary.

The third fear is the one closest to my credibility and therefore the hardest to publish.

I am not sure I know enough about AI in the African context specifically.

I know what I know. I have a financial services background, an MBA, years of watching how ideas travel between contexts and what gets lost in translation. I have a feel for when something is being copied at the surface and when someone is asking what is underneath it.

But feel is not the same as evidence. And the African AI landscape is moving in ways that are genuinely hard to track, hard to study, hard to speak about with the precision I hold myself to.

So I hold myself to something else instead. I say what I can see. I state what I do not know. I ask the questions before I deliver the answers. I stay close enough to the ground that when something shifts I feel it before it becomes a headline.

Whether that is enough is what I am still working out.

The fourth fear is one I share with everyone who does this work, but almost nobody says out loud.

The jobs.

I know the official position. AI augments, it does not replace. The people who learn to use the tools will outlast the people who do not. New roles will emerge. History shows technology has always created more than it destroyed.

I have made versions of this argument myself in rooms full of people looking at me with a specific anxiety in their eyes.

But I also know that history is not uniform in how it distributes that outcome. The new jobs do not always arrive in the same place the old ones disappeared. In economies that are already fragile, where the informal sector is enormous and the safety nets are thin, the gap between disruption and adaptation is not an inconvenience. It is a crisis.

I do not have a clean answer. I do not think anyone does. What I have is a refusal to pretend the question is settled.

The fifth fear is the one nobody wants to raise in rooms where everyone is excited about what we are building.

Data centres consume enormous amounts of energy. The tools we are all using and recommending are running on infrastructure with a climate cost that is not being addressed with anything close to the urgency it deserves.

The conversation exists. It is not loud enough. And the people who are loudest about AI adoption are often quietest about this.

I do not exempt myself. I use these tools every day. I recommend them. I benefit from them.

I just think we should say so.

So why publish this?

I have been thinking about my maternal grandfather lately. My mother does not know much about him. There is nowhere to go to find out who he was, how he thought, what his life felt like from the inside. He passed and left no record of his interior world. Just the outline of a life without the texture of it.

I woke up one day and realised that nobody knew me either. Absolutely nobody. And I decided I was not going to leave the same absence behind.

This blog is not thought leadership strategy.

It is not content marketing.

It is a woman writing herself into existence so that the people who come after her have somewhere to go.

Publishing these fears is part of that.

Not because I am comfortable with the exposure. Because I cannot write about the cost of invisible information, about the gap between what organisations say they are doing and what is actually happening, about the importance of asking the real question underneath the surface one, and then quietly protect myself from the same scrutiny.

That would not be excellent.

And anything with my name on it has to be excellent.

My own.

Thought leadership that costs nothing is not thought leadership. It is just content.