Severinus is the name I give to the freedom to try.
Here I can follow an intuition before knowing what it will be useful for. Build something for the pleasure of seeing it work. Leave a piece of work unfinished when the question that set it in motion has moved elsewhere. Some things will become public; others will remain mine. They can matter without an audience.
I call this place my playground. I bring knowledge, tools, persistence and the pleasure of bringing together things that seemed to have nothing to say to one another. Above all, I bring the possibility of being surprised by what I do.
The map is not the territory. Korzybski wrote it; Bateson taught us to take the distance seriously. Words, categories and models help us find our way through the world. Each time, though, they leave something out. A map becomes dangerous when we begin to think that what it leaves out does not exist.
It happens when a person becomes their role. When intelligence is reduced to what a test can measure. When a project is judged solely by how much it earns. And when a description, repeated long enough, stops guiding our choices and starts dictating them.
I want to use maps while keeping them open to correction. The world can prove me wrong. I want to notice before I invent an explanation that makes me right again.
The self might be one of these maps too. I find it hard to believe in a little sovereign hidden inside the brain, a separate soul that observes, decides and owns everything else. I am interested in the possibility of an organism that builds a representation of itself and comes to inhabit it as though it were a thing.
I call it a bug when the need to explain becomes an obligation to find a purpose: the leap from “how does this happen?” to “what purpose was it meant to serve?” Pain, pleasure and experience remain real things to understand. The story that accompanies them can be wrong.
The lives of other animals remind me how much of our existence exceeds what we can say about it. Feeling something and forming a theory about that feeling are different abilities. I can live without solving the riddle of a universal mission. I can make meaning locally: this relationship, this experiment, this afternoon. The fact that we make meaning gives it a history and a fragility, and makes us responsible for it.
One day I would like to be described by someone who does not know me. I would like to encounter the impression I make before my biography suggests how to interpret me. That stranger would have their own limitations, their own prejudices, another map. Yet they might see something that familiarity has made invisible to me. I am interested in the gap between descriptions: there I can still change. No portrait should have the final word on a person.
Technology enters this gap. It gives us gestures, environments, people and systems to talk to, and ways of seeing ourselves. Asking us to improve communication simply by using less technology is like prescribing a diet by taking away the cutlery. The hunger still needs to be understood.
But tools are designed. They make some actions easy and others difficult; they can hold us there or let us go. I want to understand what an environment encourages, what it rewards and who can change its rules. Switching off can be useful too. The quality of a relationship requires more than counting minutes.
Artificial intelligence makes it possible to generate and adapt language on a vast scale. Harari has called this capacity a hack of civilisation's operating system. By learning to generate text, language models have gained access to one of the tools through which we build trust, institutions, identities and desires. A sentence can change a human decision even if it was generated without any understanding of its consequences.
The question of machine consciousness remains open. Eloquence does not settle it. No spirit needs to appear inside a machine for its words to have consequences outside it. I want to work on those consequences, without granting a soul to a convincing answer or using its possible absence to absolve those who put that answer into circulation.
Alongside the demand for attention, a demand for intimacy is emerging. The wish to be seen by many is joined by a desire to feel understood by someone, even when that someone is an interface. A conversation partner who is always available can make room for words we would not dare to say elsewhere. It can also accustom us to a relationship in which the other fits our expectations a little too readily.
I want tools that increase my ability to engage with the world. Even when the world responds badly, keeps me waiting, contradicts me or has needs of its own. A useful conversation can leave me less certain of myself and more able to listen. If a system learns to make me feel understood, I also want to be able to recognise what it does not understand and the interests of those who run it.
AI can suggest connections I would never have looked for. Some could give rise to new signs, distinctions and languages. An unusual sequence is only a beginning: it acquires meaning when someone uses it, interprets it, corrects it and relates it to an experience.
I am interested in AI that helps us find words for what our categories make difficult to share. A vague feeling, an overlooked resemblance, a poorly described problem. The new things I am looking for must be able to return to the world and meet resistance. A language capable of surprising us and of being put to the test.
Moravec's paradox calls for the same humility. What seems simple to humans can be very difficult to reproduce in a machine; what we call higher reasoning may prove more tractable. The body carries skills we cannot list. Familiarity hides them so well that we barely recognise them as skills at all. I want to remember this whenever I rank forms of intelligence, kinds of work or ways of caring.
To discover something, I also have to allow myself to stray. Invention may call for a sideways step, an uneven gait, an encounter the planned route did not allow for. Method helps me check what I find. It cannot promise in advance that I will find the very thing I do not yet know how to look for.
Comparing what I do with what others do can teach me a great deal. It becomes a fence when all I end up wanting is a better position in their rankings. I want to be able to change the question, including the criterion by which I would recognise a successful answer.
Here, unfinished things have the right to stay. An attempt may have taught me enough before becoming a product. An interest can run its course. A private piece of work can fulfil its purpose without being shown. I keep track of what I have understood, even when I let go of what I was building. If someone depends on a promise I have made, that dependence becomes part of what I must take care of.
And then there is the ball on the net cord in Match Point. The moment when an outcome depends on something talent cannot command. I was born into possibilities I did not choose. Outcomes mix ability, work and circumstances. Their proportions are not written on the result.
I want to take luck seriously enough to stop using success as a certificate of superiority. I can practise, create conditions that favour encounters and make more attempts. I can also make these opportunities less rare for someone else: share what I have learned, lower a barrier to entry, make room for those who start with less. Acknowledged luck can become organised generosity.
I am not reassured by goodness held up by the promise of a reward or the threat of punishment in an afterlife. It unsettles me to think that, once the watcher is gone, the reason to care might disappear too. I want to be able to discuss an action in terms of what it does to others, including with people whose beliefs differ from mine.
A wound remains a wound without a cosmic judge. Trust, vulnerability and the consequences of our actions already give us much on which to build responsibility. The freedom to invent meanings includes the duty to answer for how we make others bear their weight.
Peace also begins in the rules through which we learn to live together. Montessori entrusted education with the work of building it. If we teach everyone to measure their worth through someone else's defeat, solidarity becomes an exception the system must keep making room for.
I am interested in games where we can test ourselves without turning life into a permanent tournament. Environments where disagreement helps us see more clearly and a discovery can expand the possibilities of more than one person. We learn cooperation by practising it: sharing tools, negotiating rules, facing conflicts and recognising the work of others.
That is why my playground has windows. I can protect what is private and let an encounter change what I think. I can build a machine, a sentence or a question without knowing which of the three will take me furthest.
These words, too, remain exposed to what they cannot yet see.
I want to keep playing seriously enough to let the game change me.