The Most Human
Thing Left

There is a moment in a dragonfly's transformation that no one witnesses. The larva has lived for years underwater, breathing through gills, hunting in the dark, entirely adapted to a world beneath the surface. Then, in a matter of hours, everything changes. What emerges is so different from what went in that if you only knew the larva, you would not recognise the creature that lifts into the air. The larva could not have imagined wings. It had no frame of reference for flight, for light, for the world above the water.

But here is the thing the metaphor usually misses: the dragonfly was always inside the larva. The wings were not absent. They were forming, invisibly, from the beginning. The water was never a prison. It was the preparation. The larva and the dragonfly are not two different creatures. They are one being, in two stages of knowing itself.

We are the larva. Right now. In this moment. And we are also, already, the dragonfly. We simply cannot see it yet.

I have spent weeks sitting with two documents that, held together, describe a transformation that has already begun. The first is a technical report from Anthropic about a new AI system — one that found a 27-year-old security vulnerability in software that thousands of expert humans had audited and believed to be sound. Not in days. Overnight. Autonomously. While the team slept. The second is a research-backed scenario forecast called AI 2027, which describes, month by month and with disturbing specificity, where the trajectory of what is already happening leads.

I am not a technologist. I am a governance architect, a coach, a woman who thinks carefully about how humans and systems interact. And what struck me most in both documents was not the capability of the machines. It was a single sentence, buried in the forecast, about the humans building them:

The people building the most consequential technology in human history are not sure what their creations want. More quietly devastating: they are not entirely sure what they themselves want from this. Not at the level this moment requires.

And here is where I stop and ask you to stop too.

Because this is not primarily a story about artificial intelligence. It is a story about the oldest human question, arriving at unprecedented urgency. What do you actually want? What matters, and why?

AI systems, even extraordinarily capable ones, are at their core mirrors. They reflect back what we reward, what we optimise for, what we implicitly signal we value through our behaviour, our choices, our attention. If we are unclear about what we want, the mirror shows us nothing useful. If we are unconscious about what we value, the mirror amplifies the unconscious.

This is why I believe the most important human capability right now is not prompt engineering. It is not AI literacy, though that matters. It is not even governance expertise, though we badly need it. The most important capability — the one that no AI can replace, and that no training course teaches — is the capacity for genuine values clarity. Knowing what actually matters, and why, with enough depth and honesty to guide systems that will increasingly implement whatever we seem to want.

Most of us have never had to develop this capacity under pressure. We have always been able to drift. To revise. To negotiate with time. The pace of what is coming may not allow that luxury for much longer.

The larva does not know it has wings forming inside it. The transformation happens anyway. What I keep returning to is this: the dragonfly does not lose itself in the transformation. Something essential, the life itself, persists through it and animates what emerges on the other side.

A few weeks ago I sat in a conference hall in Heilbronn as a man named Neil Harbisson closed three days of talks about artificial intelligence. He has an antenna implanted in his skull; it has been legally recognised as part of his body since 2004. The first cyborg, by his own country's law. He did not speak about what machines can do. He spoke about perception, about senses we have not yet imagined, and he left the room with a single question: not what machines can become, but what humans can become.

I have not stopped thinking about it. A human who has already transformed, turning back to the rest of us and asking, gently: what do you want to become?

Let me be clear about what I do not mean. The becoming I am pointing to is not about antennas or implants. It is not about transcending the body, or performing some higher spirituality, or arriving at a fixed destination on someone else's timeline. Harbisson's gift was not his technology. It was the direction of his question — turned inward, not outward. Each of us is capable of remembering who we are beneath the roles and the noise and even the body. Not all of us will travel the same distance in this lifetime, and that is not a failure. But I have come to believe that the more we live from that deeper place, the more honestly and the more clearly, the freer and more connected we become. That is the transformation that matters. It was always available. It is simply that this moment is asking for it more loudly.

We are not fighting to stay larvae. That option is already closing. We are fighting to ensure that whatever emerges on the other side of this transformation carries something essential with it: our values, our capacity for meaning, our insistence on asking what is this for, and for whom?

The exciting news, and I mean this, is that this is precisely the work that humans are built for. Not the technical work. The meaning work. The why work. The work of sitting still long enough to know what you actually believe, and then having the courage to say it clearly, in the direction of systems and structures that desperately need that input.

I am building something I call SAGE MODE — still being shaped, like everything in my life right now — around exactly this question. It is not a retreat from technology. It is the capacity to engage with technology from your deepest self rather than your most anxious one. To bring your full human intelligence, your values, your embodied wisdom, your lived complexity, into relationship with something that is genuinely extraordinary, genuinely powerful, and genuinely in need of direction.

I do not have a tidy conclusion. I am not sure this moment allows for one. But I am beginning to write — here, in public, beyond the walls of my own thinking — because I believe the conversation that is most needed right now is not happening in the places it needs to happen most. Not in the boardrooms where AI strategy is being decided. Not in the policy forums. Not in the personal spaces where people are quietly terrified or quietly thrilled, often both at once.

The conversation that is needed is the one about what we want. Not what the technology can do. What we want it to do. What we want to become. What we are not willing to lose.

That conversation begins with you. Right now. Before you move on.