The Future I Keep Predicting
What eighteen forecasts revealed about the worldview beneath more than a decade of foresight
Tracey Follows | Futurist
On BBC Radio 4's Dough, Greg Foot has asked me to predict the future of six apparently unrelated things: pets, gyms, eyewear, mobile phones, clothing and home delivery. Each episode ends with the same question. What might we expect in five years, in ten, in fifty?
Eighteen predictions across six subjects. All very different, or so I thought. When I put them side by side though I realised I had not made eighteen independent forecasts. I had been telling variations of the same story. And when I went back through more than a decade of published forecasts, research and essays, I found I had been telling it for rather longer than I knew.
(The predictions themselves are listed at the end.)
The story underneath
It begins with things disappearing. The phone matters less as an object and more as computational infrastructure, until, at fifty years, it is less a screen than an invisible brain behind a personal AI. Eyewear becomes a layer through which the world is interpreted, until sight itself becomes software. Clothing becomes a surface for projected identities. The home becomes part of the logistics network. In each case the technology does not vanish, it dissolves into the environment.
I recognised this from much earlier. In August 2015, I published Rethinking the Internet of Things, proposing three phases in the evolution of connected technology: the Internet of Things, defined by relationships between devices; the Internet of Un-things, defined by relationships between data as products and services became increasingly dematerialised; and finally the Internet of Thoughts, defined by relationships between mind and body as computing moved towards biological and cognitive interfaces.
I was writing about advertising and communication, but the argument had already moved beyond devices. I imagined computers migrating from people's pockets to their clothes and eventually into their minds. I suggested that purchases might one day be made by thought rather than gesture, and that brands could anticipate desires before customers had consciously expressed them. The distinction I made was between systems designed to assist and systems capable of anticipating. Eleven years later, the domains have changed considerably. The underlying assumption has not.
In January 2022, I made a similar argument on The Megyn Kelly Show, describing a move towards ambient and spatial computing, where devices recede into the background and interaction becomes part of the environment. I talked about augmented-reality contact lenses, connected clothing and gestures replacing interactions with a smartphone. When Megyn asked about gaming, I pointed out that forms of embodied interaction were already happening in virtual reality, including body tracking and biofeedback. Once again, the underlying forecast was not the arrival of another device, but the gradual disappearance of the device as the primary means of interacting with technology.
From assistance to anticipation
In the near term, my forecasts describe AI assisting us: analysing fitness data, enhancing vision, recommending what to wear. Further out, assistance becomes anticipation. The system no longer waits for me to formulate a request. It interprets my behaviour, infers my needs and increasingly acts upon them.
Home delivery shows the progression most clearly. At five years, the platform scores me as a customer. At ten, my home has been rebuilt around its logistics. At fifty, it may not need me to place an order at all.
In a longer version of that scenario, I imagined a neural commerce cloud, in which AI agents and brain-computer interfaces interpret our desire signatures and negotiate on our behalf. a world in which mood, emotion and subconscious inclination become commercial signals. Not unconnected from the Sky Future of Media predictions I made in 2019 that included ‘empathetic media’.
This remains a speculative fifty-year scenario, not a claim about what today's brain-computer interfaces can reliably determine. But it illustrates a recurring movement in my forecasts. A technology designed to satisfy our needs becomes one that helps decide what those needs are. The question is no longer whether delivery can become frictionless, but whether desire remains ours to formulate.
Personalisation becomes individual reality
I rarely imagine one technological future experienced uniformly by everybody. I imagine increasingly individual ones. Fitness is designed around a biological and cognitive profile. Glasses interpret the same street differently for each wearer. Clothing appears differently to different observers. AI agents arrange information, services and opportunities according to what they believe they know about us.
Taken to its conclusion, personalisation stops being a matter of content and becomes a matter of reality. Two people can stand in the same physical place and inhabit different informational, commercial and perceptual worlds.
In 2019, my independent report for Sky, The New Dimensions of Media, explored a similar movement towards smart environments and empathetic media: systems capable of sensing, interpreting and responding to people's emotional states.
By the time I wrote The Future of You in 2021, I was asking what happened when these systems began constructing representations of the person herself. If everything around us can be personalised, and personalisation depends upon a model of who we are, then the accuracy, authority and influence of that model become consequential.
Personalisation is no longer simply about getting what we want. It begins to shape what we experience, what we encounter and potentially what we believe is possible. Which raises a question my forecasts keep returning to: what remains of the common world we still have to share?
From the augmented body to the shared mind
Running through the predictions is another persistent assumption: data moves progressively closer to the self. Technology begins by observing what we do, then measures what our bodies are doing, then interprets what we see, mediates how we appear and eventually attempts to infer what we want and think.
That movement from behaviour to embodiment, perception, cognition and desire runs throughout my TRACE Foresight Timeline.
In 2016, I wrote about augmented athletes and how human competition might be transformed through biological and technological enhancement. But the clearest examples came in 2021. In The Shared Mind, I explored the possibility of brain-computer interfaces enabling connected cognition, and what individual creativity might mean if minds became more directly networked. In my BBC Radio 4 Four Thought talk, Virtually Immortal, I considered digital afterlives: technological representations of people persisting beyond biological death.
These two subjects reappear in the Dough predictions. The fifty-year pet forecast imagines an AI representation of a dead animal continuing to inhabit the home. The fifty-year delivery forecast imagines brain-computer interfaces connecting thought and commerce. The gym forecast, meanwhile, imagines resilience becoming a measure of physical, mental and emotional capacity, while eyewear evolves towards the interpretation of perception itself.
Very different subjects, but one recurring transformation. More of the world, and more of the person, becomes legible to technological systems. And what becomes legible can be modelled, predicted and acted upon.
When a prediction becomes a condition
Somewhere in that progression, the forecasts stop being about capability and start being about power.
Take biological age. In the gym, it is useful information about health. In my fifty-year scenario, it has become a resilience or longevity score that insurers, employers and pension providers might consult. The technology no longer merely helps someone understand themselves. It becomes part of the infrastructure through which others evaluate them.
Home delivery follows the same path. Rewards for efficient behaviour become customer scores, and scores make access conditional. We begin with a convenience and end with a condition of participation.
This concern appears in my earlier work beyond consumer technology. In 2017, I contributed to a Future Visions 2035 discussion for Virgin and highlighted predictive policing as the trend I most wanted to predict, exploring the ethical questions raised when algorithms anticipate behaviour. The central issue was not simply whether a system could calculate risk, but what happens when its predictions influence decisions about real people.
There is a world of difference between a system describing who I am and a system determining what I am allowed to do. The machine-readable version of the person becomes consequential for the lived person. That, I think, is one of the deepest assumptions in my work: the most significant technological changes eventually become changes in the relationships between people, institutions and systems.
The shape of fifty years
The time horizons have a shape of their own. At five years, I tend to imagine augmentation: familiar practices become more capable. At ten years, reconfiguration: objects, systems and environments change their roles. At fifty years, the forecasts become ontological. They ask what something is, or what a familiar human activity might become.
What is a pet whose digital presence outlives it? What is ageing when biological measures displace chronological ones? What counts as seeing when machines interpret the visual world? What is clothing when appearance becomes less dependent on physical material? And what counts as wanting when desire can be inferred before it is consciously expressed?
That is why my longest-range forecasts often sound less like predictions than thought experiments. I use the far horizon to ask what happens when a familiar distinction can no longer be taken for granted.
What the radio leaves out
A consistent pattern is not proof that the pattern is right. Read on their own, these forecasts might suggest a futurist who sees only one direction of travel: integration, acceleration, ever-closer machines.
That would misread them. A broadcast prediction is a compressed thing. With so little time for each horizon, I describe the trajectory, because the trajectory is what the question asks for. What there is rarely time to say is that every trajectory meets a countertrend, and that the countertrends are often where the future is actually decided.
This is the heart of System & Self. Systems propose; selves respond.
People do not take technological change wholesale. They push back, protest and refuse. They also engage, adapt, augment and personalise, bending a technology to purposes its designers never intended. They contextualise it within their own lives, cultures and relationships. Societies legalise some things and prohibit others, and then change their minds. The future is not simply what the system delivers. It is what emerges from the negotiation between the system and the self.
My fifty-year delivery scenario already contained one such response: a counterculture that deliberately resists anticipatory commerce, choosing delay, ritual, uncertainty and the experience of wanting something without having it at once. Its members might prefer to stay hungry, to miss out, or to make a decision for themselves rather than have it made for them.
In a world where everything can be anticipated, the unanticipated may command a premium. The ability to wait, to wonder, to change one's mind or to refuse becomes a kind of freedom. So the closer machines move towards the self, the more the boundaries of the self matter. That is not the opposite of my forecasts. It is the other half of them.
Who sees which future?
There is a further element the forecasts cannot show on their own: who is doing the seeing.
Every futurist looks towards the future from a particular point in historical time. I make my forecasts from my own generational vantage point, having come to networked technology from a world of broadcast media, physical products and relatively discrete communication devices. I recognise the movement from objects to environments as a transformation because I remember what those objects were before they became connected or computational.
Someone who has grown up with voice assistants, recommendation feeds and AI companions may experience those developments differently. Smart glasses that feel to me like an overlay on the world may feel to them like part of the world. An anticipatory system I regard as potentially intrusive may feel to them like being understood.
But generation is only one influence. The media environment is another, and the two should not be confused.
Marshall McLuhan argued that the medium is the message: that media environments shape how we perceive and understand the world, not merely which information we receive. Broadcast media, algorithmic feeds, immersive interfaces and AI-generated environments organise attention, perception and participation in different ways. People of the same generation can inhabit radically different media environments, just as people from different generations can inhabit the same environment but interpret it differently.
Generational experience shapes the historical sequence through which we encounter technological change. Media environments shape the conditions through which we perceive it. Culture, geography, institutions and individual circumstances exert their own influences too.
This matters for forecasting because the same technological development may produce different experiences and social consequences depending on who encounters it, and under what conditions. What I perceive as an entirely new relationship between humans and machines may be experienced elsewhere as an ordinary extension of existing habits.
The future is not one reality arriving for everyone at once. It is many realities, encountered through different histories, cultures and media environments, by people whose understanding of the world has been formed under different conditions.
Every futurist has priors
Every futurist carries hidden assumptions about how change works. Some expect disruption, others continuity. Some see cycles, others accelerating curves. Some begin with economics or geopolitics, others with technology, culture or power. Give us enough unrelated things to forecast and, eventually, those assumptions show.
That is why I think there is value in putting one's predictions side by side every so often. Not only to count which were right, though that matters, but to examine the model of the world they came from. A forecast is never only a statement about what might happen. It is also a statement about what its author believes causes change.
The world I keep sketching is one in which systems grow ever more capable of interpreting and reconstructing reality, including the people within it. Objects become interfaces; interfaces become environments; information becomes interpretation; interpretation becomes prediction; and prediction becomes a basis for action.
That orientation eventually led me to System & Self and to Me:chine. I do not want to read those later ideas back into every earlier forecast, as though I had been building the same theory all along. But what I can see is a persistent interest in what happens to human experience, identity and agency when technology becomes part of the environment.
In 2015, I was asking how the internet might move from things towards thoughts. In 2026, I find myself asking what happens when systems can act on thoughts, and what room remains for mental privacy. But also for people to think, desire and imagine something else. That is not quite the same question. It is what happens when one question grows into another.
Perhaps I have not been predicting the disappearance of the human into the machine at all. Perhaps I have been predicting the growing importance of the relationship between them, and of the many different people, in many different media worlds, who will negotiate it.
Whether that is the future, or simply the future I am predisposed to see, is a question I intend to keep open.
The eighteen predictions
Pets
Five years: The definition of a pet expands to include creatures such as foxes and crows.
Ten years: Smart collars, cameras and AI analytics make animal health and behaviour increasingly measurable.
Fifty years: Rather than cloning pets, people preserve their memories through AI recreations, holograms and digital companions.
Gyms
Five years: Biological age begins to matter more than chronological age.
Ten years: Training is personalised through biological data, cognitive assessment and AI, and oriented towards enhancement as well as health.
Fifty years: Exercise becomes resilience training, incorporating mind and body, with humanoid coaches. Longevity scores may shape how institutions evaluate people.
Eyewear
Five years: Augmented glasses offer better medical assistance and AI support.
Ten years: Eyewear explains the world through translation, object recognition, context and health signals.
Fifty years: “Sight becomes software”, with information experienced as layers of interpretation within the visual world.
Mobile phones
Five years: Phones become quieter as we detach from their constant demands.
Ten years: The smartphone becomes one component in a network of wearables, glasses and connected devices.
Fifty years: Personal AI displaces the phone, which survives, if at all, as an invisible computational brain.
Clothing
Five years: Resource pressures, sustainability concerns and AI reshape what we buy and wear.
Ten years: Clothing adapts to temperature, health and environment, and virtual identities acquire their own wardrobes.
Fifty years: Clothing may be streamed rather than physically worn: one material garment, many augmented appearances for different observers.
Home delivery
Five years: Social Credit Delivery, in which platforms reward or penalise customers for their behaviour.
Ten years: The home becomes part of the logistics network, with purpose-built infrastructure and potentially humanoid robots.
Fifty years: Telepathy as Delivery, in which brain-computer interfaces may let systems infer what we want before we express it.
These are condensed summaries of my predictions on BBC Radio 4's Dough. They are possible futures rather than statements of technological inevitability. The longer history of my published forecasts and research is available in my TRACE Foresight Timeline.