Four Conditions of Futurity

Authored, Outsourced, Invisible, Imposed

Futures thinking has a rich vocabulary for distinguishing between kinds of future. One of the best-known ways of distinguishing futures is through the familiar language of possible, plausible, probable and preferable.

These distinctions matter because they help us resist the idea that only one future is waiting to arrive. They invite us to consider alternatives, to separate likelihood from desirability, and to recognise that what looks probable is not necessarily what we should want.

I have no argument with any of that. Quite the opposite. But I have come to think there is another question we need to ask. It is not another category of future, or another cone, matrix or set of scenarios. It is a question about agency.

Before, during or after we decide whether a future is possible, plausible, probable or preferable, something else remains to be understood:

What is our relationship with that future?

Did we take part in imagining it? Has somebody else imagined it for us? Can we even see the future being constructed around us? Or has it already started to determine what is possible?

I think of these as four states of agency, producing four conditions of futurity:

Authored. Outsourced. Invisible. Imposed.

They are not alternatives to the familiar categories. A possible future might be authored or imposed. A preferable future might be outsourced. A probable future might be invisible to the person whose life it is already shaping. The two vocabularies answer different questions. One asks what sort of future is this? The other asks:

What is the state of my agency in relation to it?

The future is no longer only ahead of us

This question matters because the future has changed its position in our lives.

We still think of the future as something ahead of us, something we imagine, forecast or prepare for. But futures are now also being produced around us, in the present.

A bank holds a model of whether someone is likely to repay a loan. An insurer holds expectations about future risk. A recommender system predicts what someone may want next. A security system assesses who is likely to pose a threat, and a health system anticipates future illness.

None of these futures has happened. Yet they already have consequences. A predicted future can influence a price, a ranking, an opportunity, an intervention or a decision.

Futures are no longer simply things we look to. They are things that look back.

That is why agency now matters so much. It is one thing to ask which futures might happen. It is another to ask what happens when somebody else's version of the future starts acting on you before that future exists.

Here are my four differently conditioned futures:

Authored futures

The first condition is the authored future.

Authorship does not mean control. No individual controls what comes next, not entirely. Systems intervene, other people can intervene, sometimes fate intervenes. Any serious account of the future has to begin with uncertainty.

Authorship is about something more modest. It means I still recognise myself as a participant in what comes next. I can imagine alternatives and exercise judgement. I can make choices and revise them as circumstances change. I can refuse one direction and maybe pursue another. The future stays open enough for my desires and intentions to matter.

Authorship is also different from participation. A person can be invited into a foresight exercise without gaining much authorship over the future under discussion - being included in somebody else's process is not the same thing. Authorship concerns whether I keep a meaningful relationship with my own futurity. Can I still make something of what comes next?

That is the state of agency I am interested in.

Outsourced futures

The second condition is the outsourced future.

There is nothing inherently wrong with outsourcing anticipation. We all do it. Governments commission forecasts, organisations hire strategists and futurists, individuals rely on experts, markets aggregate expectations, and models help us understand complexity.

But when we stop noticing, the imaginative work has moved somewhere else. An expert presents the forecast or a model sets the probabilities, an AI system generates ten possible futures in seconds. The future, in these cases, comes ready-made.

We may still be free to choose among the options presented, but the underlying act of imagination has already happened. Someone else framed the question. Someone else decided which variables mattered. Someone else drew the edges of the future.

Generative AI makes this especially interesting. The apparent abundance is extraordinary. Ask for five futures or scenarios and you get five; ask for fifty and you get fifty. But more futures do not necessarily mean more futurity. If the same assumptions sit beneath every scenario, the future can multiply without becoming more open. Fifty machine-generated futures are not necessarily more open than five human ones.

And now futurity itself is being outsourced to the platforms. Anthropic has built an interactive model that can generate alternative economic futures for 2030, combining assumptions about AI capability and adoption with jobs, wages, unemployment and growth. OpenAI has gone further still, establishing its own Strategic Futures team to investigate how transformative AI might reshape firms, institutions and society. These companies possess computational power, modelling capacity and machine intelligence at a scale no individual could hope to reproduce.

The future can therefore arrive pre-modelled: scenarios generated, consequences calculated, possibilities laid out before us without our having lifted a finger or bothered a brain cell. We no longer have to imagine our way into the future, we can just ask the system to imagine it for us.

The outsourced future can therefore create an illusion of plurality. At the same time we are offered more futures, we are doing less of the imagining ourselves.

Invisible futures

The third condition is the invisible future.

Here the issue is no longer that somebody else has done the anticipating. We may not even know that anticipation is taking place.

We now inhabit systems that contain models of what is likely to happen next, and those models may never be visible to us. We may not see the score. We may not know which behavioural signals produced an inference, or which future version of ourselves the system has generated.

We experience only the consequence: a price, a recommendation, an extra check, an opportunity surfaced, another withheld.

The future hasn’t disappeared. Actually, quite the opposite. The future is inside the system - but it is opaque. And that creates a particular problem for agency, because:

You cannot contest a future you cannot see.

Traditional foresight tends to make futures explicit. We put scenarios on the wall, name assumptions and debate uncertainties, precisely so they can be discussed. Invisible futurity reverses that logic. Its anticipatory assumptions become powerful because they disappear into the operation of the system.

There is another kind of invisibility too. Mark Fisher wrote of a culture in which the existing order had become so dominant that coherent alternatives were increasingly difficult to imagine. The problem was not that the future had vanished, but that certain futures had moved beyond the horizon of what could be conceived as realistic. Predictive systems can intensify that condition. By continually presenting what is likely, relevant or optimised from what already exists, they can make alternative trajectories progressively harder to see. A future does not have to be prohibited to disappear; sometimes it simply falls outside the field of vision. And what cannot be seen becomes difficult to imagine, difficult to choose and ultimately difficult to author.

It is the same logic behind the familiar phrase, “if you can’t see it, you can’t be it”: what never becomes visible as a possibility can scarcely become imaginable as a future.

This is the condition in which possible futures are squeezed from view before they can be imagined: by never becoming visible as possibilities, they are denied the chance to be authored into being.

Imposed futures

The fourth condition is the imposed future. This is where too much anticipation flips over into an imposed future.

If a lender believes someone is unlikely to repay, that imagined future affects their access to money in the now. If an institution expects a place to generate certain kinds of risk, that expectation can change how attention and resources are distributed. If a platform predicts that I prefer only one kind of information, that can narrow the way my preferences develop.

The future starts organising the present. Over organising it, in fact.

None of this requires the prediction to be wrong. The issue is whether the person affected keeps any meaningful ability to challenge the future assigned to them under the weight of the prediction.

Sociology has long recognised this through the idea of the self-fulfilling prophecy: once an expectation is treated as real, people and institutions may begin behaving in ways that help to make it real. More recently, researchers have described something similar as performative prediction, where a prediction influences the very outcome it purports merely to anticipate. The heavier the prediction becomes, through repetition, quantification, institutional authority or sheer familiarity, the more easily possibility begins to feel like fate. At that point prediction does not simply describe what may happen. It starts organising behaviour around the assumption that it will. The future is pre-empted precisely because it has been made to feel predetermined.

This is where futurity becomes inseparable from power.

Four conditions, not four boxes

Authored, Outsourced, Invisible and Imposed are not four tidy boxes into which every future must fit. A situation can contain several at once. I might author one part of my future while outsourcing another. A future generated elsewhere might be visible at first and later become embedded invisibly. An outsourced future might stay advisory, or it might slowly become imposed.

They do, though, describe a rough gradient.

At one end, I make the future.

Then someone makes it for me.

Then I cannot see it being made.

Finally, I must live inside what has been made.

Agency diminishes along that line, and responsibility becomes harder to locate. Who authors an outsourced future? Who is accountable for an invisible one? Who decides when a prediction carries enough authority to become a condition?

The point is not classification for its own sake. It is to investigate the state of agency.

How much authorship remains? What have I outsourced, and what can I see? What is acting on me, and what can I still contest?

Those questions reveal something likelihood alone cannot. A future can be highly preferable and still be imposed. It can be probable and still be invisible. It can be possible and yet imagined entirely by somebody else. And a future can be deeply uncertain while remaining strongly authored.

A future can be good for me and still not be mine.

So the two vocabularies describe different things. Possible, plausible, probable and preferable help us think about the character of futures. Authored, Outsourced, Invisible and Imposed help us think about the condition of agency around them.

The first maps possibility. The second maps power.

System & Self

This distinction has emerged through my work on System & Self, which starts from a simple proposition:

Every future is written between the system and the self.

The system is the external environment: technologies, institutions, infrastructures, economics, rules, networks and power. The self is the lived experience of those systems: identity, judgement, agency, behaviour, meaning and adaptation.

The point is not to choose between them. The important territory is the relationship. What is the system doing to the self? What can the self do back? What happens between systems? What happens within the self as it responds?

Seen this way, agency becomes much less abstract. The self appears not as a demographic or a stakeholder, but as a person standing in relationship with systems that are constantly making claims about what comes next. Different futures place that person in very different relationships with the systems producing them. Some invite authorship. Some encourage us to outsource. Some operate invisibly. Some begin to impose.

That is why I think the four conditions belong naturally within System & Self.

They give the question of agency greater specificity. Rather than asking: do I have agency?, we can ask what is the current state of my agency?

That is a far more useful question.

Half an Hour of Future

It also sharpens a practice I have been developing called Half an Hour of Future.

The idea is deliberately small. Much contemporary futurity is experienced at enormous scale: AI transforming civilisation, climate breakdown, the future of humanity. These questions matter, but their scale can leave the individual with little sense of where to stand.

Half an Hour of Future does the opposite. Using System & Self cards, I take one tension, one relationship or one question and spend half an hour with it. The purpose is not to predict. It is to surface something that might otherwise stay submerged: an assumption, a possibility, a question of agency.

(you can stay up to date with the live readings here)

The four conditions make the practice more precise. Once a System & Self reading has surfaced an issue, we can ask: Is this a future I am helping to author? What part of it have I outsourced? What remains invisible? Where might expectation already be turning into imposition?

That does not turn Half an Hour of Future into another elaborate framework. Quite the opposite. It gives the practice a clearer question:

What is the state of agency here?

That may be enough.

From Too Much Future to a State of Agency

This connects to something else I have been writing about: Too Much Future.

By Too Much Future I do not simply mean too many forecasts. I mean that we are becoming saturated by futurity generated elsewhere. Scenarios, predictions, risk assessments, simulations and machine-generated possibilities surround us. Some are useful, some necessary. But quantity alone does not produce agency. There may even be a point at which more anticipation produces less.

We can have thousands of forecasts and very little say in the assumptions beneath them. We can be surrounded by predictions and find it harder to imagine anything different. We can have more future and less authorship.

That is why the four conditions matter. Too Much Future can be outsourced. It can be invisible. It can become imposed. The question then is how we recover the possibility of something more authored.

That does not mean total control, or the fantasy that individuals can choose whatever future they like. It means enough agency to remain participants in what comes next.

A different question for foresight

Foresight will always need to ask which futures are possible, and to investigate plausibility, probability and preference. Those questions remain valuable. But the spread of predictive systems means another question now sits alongside them. It is not about another future, but about our relationship to the future.

Who is doing the imagining?
Where has anticipation been outsourced?
Which futures are operating invisibly?
When does prediction become imposition?

And beneath all of them: what is the state of our agency?

This shifts foresight from asking merely what might happen? to asking what is this future doing to us now?

As AI expands the scale and reach of prediction, the future is no longer simply something we are trying to see. Predictive systems are already constructing futures about us: scoring what we are likely to do, need, want or become, and allowing those anticipated futures to shape decisions in the present.

So the task is not only to widen the range of futures we can imagine. It is to preserve the conditions under which we can still take part in making them.

Authored. Outsourced. Invisible. Imposed.

Four states of agency. Four conditions of futurity.

And perhaps a more precise way of asking the question that sits beneath all foresight:

Who still gets a say in what happens next?

ps. A note on intellectual lineage

These ideas sit alongside a wider body of work on anticipation, prediction and power, including research on anticipatory action, algorithmic governmentality, behavioural futures and performative prediction. Writers such as Ben Anderson, Antoinette Rouvroy, Shoshana Zuboff and Elena Esposito have shown, in different ways, that predictions do not simply describe possible futures. Once embedded in systems and decisions, they can begin to organise the present. The distinction I am making here is to look at that territory through the state of the self’s agency: whether the future remains authored, has been outsourced, has become invisible, or is beginning to be imposed.