When people talk about the future of AI, they usually imagine one of two things: either a terrifying machine overlord, or a glorified office assistant summarising emails while slowly encouraging humanity to communicate entirely in bullet points. Neither vision inspires me much.
The future I want looks rather different.
I imagine an elderly couple in a large garden somewhere in northern Europe. Old apple trees. A vegetable patch. A greenhouse full of tomatoes that actually taste of something. A few cats sleeping in improbable places. And trotting quietly behind them, a small robot companion.
Not a humanoid servant. Not a chrome-plated butler announcing calendar appointments in a soothing Californian accent. A robot dog.
It follows you around the orchard while you inspect the apples. It notices that one tree is showing signs of fungal infection before you do. It remembers where the best mushrooms appeared last autumn. It warns you that the cider fermentation smells slightly wrong. It quietly maps the flowering times of your wild fennel over several years and notices subtle changes in the climate before any government report does.
And because it is connected to an advanced language model, you can talk to it naturally.
“Have the pears by the north hedge ripened yet?”
“The southernmost tree is ready. The others need another week.”
“Do these leaves look diseased?”
“Probably magnesium deficiency. Similar symptoms appeared after the wet summer two years ago.”
“Where did we find the good chanterelles last October?”
“Near the old stone wall after three dry days followed by rain.”
That, to me, sounds like civilisation improving itself.
The strange thing is that this future is technologically far more plausible than many of the grand promises we hear. We already have language models, cameras, microphones, soil sensors, GPS systems and increasingly capable robotics. What is missing is not really the technology. It is the vision.
Much of Silicon Valley still imagines the purpose of AI as replacing humans in offices. Endless presentations discuss “productivity”, “disruption” and “automation pipelines”. The machine is imagined primarily as a cost-cutting device.
But perhaps the real promise of AI lies elsewhere.
Perhaps its greatest contribution will not be replacing human life, but helping us engage more deeply with the physical world again.
For decades, technology has often had the effect of pulling us away from reality and into abstractions. More time indoors. More time staring at screens. More time handling documents, notifications and administrative trivia. Even supposedly “smart” devices often make life feel strangely thinner.
A good AI companion could do the opposite.
It could encourage people to garden, walk, forage, observe birds, understand weather patterns and rediscover seasonal rhythms. It could help elderly people remain independent longer. It could preserve practical local knowledge that currently disappears when people die. It could help children learn the names of plants and insects instead of merely recognising corporate logos.
In other words, the AI revolution might ultimately become valuable not because it distances us from nature, but because it reconnects us with it.
The irony is delicious. After centuries of industrialisation pulling humans away from the land, the descendants of today’s large language models may end up helping people return to it.
I find that a much more hopeful vision than another app for optimising meetings.
And yes, if the future includes semi-autonomous robot dogs helping people manage orchards in rural Denmark, Scotland or Tuscany, I suspect humanity may, after many detours and absurdities, finally have started building technology for civilisation rather than merely for commerce.