Does AI Evolution Herald Human Devolution?

In 2017, BBC’s Science Focus asked “Dude, Where’s My Flying Car?” and proceeded to outline the predicted release dates of futuristic technologies long since dreamed up by people for centuries. The list features the titular flying cars, cyborgs, time machines, and robot maids.

As technology has developed, so has our thinking around it. Traditional sci-fi speculation surrounds machines that can either do things for humans or enable them to do them — fly, turn invisible, do our dishes — all while chrome-clad and with rounded edges. These machines exist outside of our bodies and are as alien and incompatible with our humanity as any other inanimate object.

In the distant future, humankind has taken technology to its extremes, achieving faster-than-light space travel and boosting our exploration of space and all that could inhabit it. And at the core of all this sits the human mind in the driver’s seat.

Perhaps this vision of futuristic technology was easier to imagine than where we have ended up. Machines are easily conceptualized, but cultural shifts and sociological development are harder.

In the last few decades, technological vision has shifted away from the machines toward that which powers them; the algorithm, artificial intelligence, and their alignment with human beings and our culture.

The machines we dreamed of always kept their distance. They hovered, they whirred, they did the dishes — but they stayed on the other side of the room, chrome and separate, things we could switch off and walk away from. The technology that actually arrived provides no such distance. It does not sit across from us; it watches us, learns us, and reaches steadily inward — first into how we behave, then into what we want, and at last toward the parts of us we cannot consciously control. The question is no longer what our machines will do for us. It is whether, by the end of it, any part of the self is left that they have not touched.

Our relationship to the camera

It’s an often quoted statistic that the average person living in Britain is caught on CCTV footage 300 times per day. A startling number that gives pause whenever one hears it. However, the statistic is from 2004. The Independent article that broke the news at the time warned of the invasive nature of Oyster cards and microchips in bank cards, and the very real threat to privacy with a government tracking purchases and movements. Concerns that today almost seem naïve, as such novelties quickly became necessities.

Orwellian forms of surveillance are becoming commonplace. Big Brother emerged in a post-9/11 world promising safety, and remains through the promise of convenience. Yes, he is watching, but only so our route to work is accurately mapped out with potential obstacles. Yes, he is watching, but only to make sure we are never inconvenienced with an advert we would not be interested in. Yes, he is watching, but it is for our safety, comfort and convenience.

He cannot operate alone. Early stages of CCTV and lower-tech surveillance placed the guard tower firmly in the centre of the Panopticon, with governments and tech-oligarchs the only ones watching. But lasting change comes from affecting the culture. Ordinary people have adopted the methods of surveillance, with household CCTV and phone cameras as first lines of defence against any perceived threat.

As soon as the cameras were in our own hands, we turned them on ourselves, and the line between guard and prisoner began to dissolve. What started as the occasional “photo dump” — a camera roll emptied out after the event — hardened into the daily documentation of ordinary life. Sociologist Erving Goffman called it the presentation of self; social media has made it a way of living, in which we perform our days rather than live them, staging the unremarkable and even manufacturing moments for an audience we assume is always there. It fills our feeds now: the “day in my life,” the “get ready with me,” the most private corners of a life offered up unprompted.

The increasingly common practice of sharing imagery from private Ring cameras — in neighbourhood Facebook groups under the guise of raising awareness of antisocial behaviour, or on social media when a delivery driver has a sweet interaction with a dog — shows how normalised it has become to film people without their knowledge or consent, and to treat that footage as ours. The discomfort of imagining ourselves on the other side of the lens is real, but we accept it as an everyday risk. And the lens is everywhere: Any passerby might end up on a stranger’s feed, either deliberately or as the backdrop to someone else’s content, pulled into a conversation they never agreed to join — a cell as small as a phone screen and as wide as the internet itself. This erosion of privacy on an interpersonal level is what paves the way for the surveillance state.

Socially, the impact of the past 20 years cannot be overstated. The so-called “cancel-culture” we live in, whatever its net worth, is a clear manifestation of internalized authority and the policing of the self and others: Even with no one around, we know what may and may not be said, much less posted. We extend the same scrutiny to each other — judging how diligently a person self-surveils, quietly distrusting those who don’t toe the line. Is the professional with no LinkedIn trustworthy? What about the date with only a handful of photos?

Younger generations, who have grown up inside the surveillance state, show the clearest signs of all. Natives of the Panopticon, their lives are as real on camera as off, and they act accordingly: The Z and Alpha generations drink and party less for fear of being caught on camera. Who needs strict parents when Big Brother and the algorithm are always watching? Even the rebellion is choreographed — the dance rehearsed, the prank on-trend.

And the grey areas are widening faster than the law can close them. At its extremes it sometimes catches up — revenge porn, for instance, is now broadly criminalized — but newer tools keep outrunning it. Consider the genre of male “pickup” influencers, who film their approaches to women using smart glasses, the recording light hidden under a sticker, and post the footage for views. Women have been turned into clips watched by millions, recorded without consent and harassed once the video spreads, with little recourse: In the UK there is no specific law against filming a stranger in public. It is the same instinct traced to its end. Surveillance came to us as protection; we took it up as a tool; and now, in hands like these, it has become a weapon.

Wants and needs, pre-determined

While we perform in the yard — filming ourselves, judging one another, glancing over our shoulder for the stranger’s lens — the prison itself, the algorithm, is quietly at deeper work. It studies us, learning each of us from everything we do and assembling us into a profile it can sort and predict. The 2017 BBC article called this “mind-reading machines” and ironically said it was 2-10 years away.

Deleuze gave this data-made version of a person a name: the “dividual.” He posited that once the algorithm can predict us, it can begin to shape us, arranging what we see until our responses turn reliable. This is what it means, in the end, to be aligned: not argued into anything, simply adjusted. We hand the system our deliberate choices, and it hands back a revised set of preferences we experience as entirely our own.

While not a prerequisite of the surveillance state, it is fuelled by capitalist ideology. By shaping a predictable individual, the most valuable form of consumer is created.

You can watch this happen to taste at scale. A generation raised on the same few recommendation engines drifts, without noticing, toward the same narrow band of likings — part of how a handful of globe-spanning names, the Ed Sheerans and Taylor Swifts, came to fill a space that once held thousands. The human experience is boiled down to Spotify daylists titled “power ballad chai latte Monday morning,” “spooky evil gay mix,” or “always single bed rotting Wednesday afternoon.”

Music used to be local: A town could keep its own musicians, most of them making a decent living without ever being famous. The algorithm dissolved the local. It pushes everyone toward the few things it already knows will perform, and in the winner-takes-all economy that follows, only the very top earns a living at all. The flattening runs both ways — it narrows what we consume and what we dare to make, every creator chasing the same viral shapes because those are the shapes that get rewarded. We are not only sorted into types. We are learning to produce ourselves to type.

What we take in exchange is convenience dressed as intimacy: The feed that knows us spares us the effort of choosing. It cultivates a new form of social capital. But that convenience is a story the system tells about itself. Our real needs were met long ago; what it manufactures now is the steady supply of new wants.

There is, though, a ceiling to sameness. With every generation there is less differentiation between personality categories, until eventually millions of people are listening to daylists with different titles but the same songs. A culture with one song left is a monolith, and a feed that has flattened everyone into the same few tastes runs out of things to sell them. So the next turn is already under way, running the opposite way — away from mass sameness toward its mirror image: not one feed for the crowd but a different one for every nervous system in it. To tune itself that finely, the system has to know us far more intimately than a click has ever told it. It has to get underneath the choices we make on purpose.

Have we taken the self for granted?

The means to achieve this is already arriving. Computing is leaving the screen and dispersing into the room and onto our bodies — glasses, rings, wearables, the smart home. It no longer waits for us to act. It reads the body directly: where the eye lingers, the catch in the breath, the rise in the pulse a half-second before we know we have reacted.

This is the threshold the earlier stages never crossed. Until now the watching stopped at what we did — the conduct we performed, the choices we made — and what we did, we could in principle withhold. The involuntary body cannot be withheld. When the prison reads the flinch before the thought, there is no longer an inside to retreat into. Freedom of thought has always rested on a small privacy — that our reactions were ours until we chose to reveal them. Ambient computing prizes that privacy away. The guard no longer infers what we are thinking; it knows before we do. Even those opting out of wearable tech will be influenced both culturally and through the data-driven options developed for them based on other people’s behaviors.

And the trade, absurdly, is smaller than ever. For the surrender of the last private room we are offered a timer that hovers above the stove, a light that knows when we enter, a slightly shorter path to the thing we were going to buy anyway. The cost climbs as the bribe shrinks.

Until now, the guard and the prisoner had collapsed into us, while the prison — the algorithm — still worked from outside, a structure we performed within and could, in principle, step away from. Ambient computing brings the prison inside. When the watching moves into the nervous system there is no external structure left to be contained by, because the walls are our own attention; tower, guard, prisoner, and prison meld together. And because the system now reads and shapes us at once, in real time, it begins to act on the self while the self is still forming — not merging with us, nothing so crude as a machine in the flesh, but co-writing us: meeting a reaction before we know we have had it, and adjusting what comes next. The self stops being something we author alone. That is the question the future forces on us: when our inner reactions are answered and shaped before they reach awareness, who is the author of the self that results — and is there still a single “we” doing the reacting at all?

Conclusion

This is where the Fermi paradox stops feeling like cosmology and starts to feel like a warning. The universe is old and vast enough that it ought to be loud with other minds, and yet we listen and hear nothing. One account of that silence is the Great Filter: a step on the road to a lasting civilization that almost nothing survives and that may lie ahead of us rather than behind. We picture it as a catastrophe — the warhead, the asteroid, the climate tipped past saving. Ours might be gentler. A civilization does not have to be destroyed to vanish; it can simply hand its thinking to its machines, one convenience at a time, until it forgets who it was as a people and lets the skills it once needed to survive wither away. Such a species would not end in catastrophe. It would just, gradually, go quiet.

And we do not have to imagine how a people fails to notice itself being remade. We have already watched it happen, in miniature and within living memory. We welcomed the smartphone as a convenience and lived inside it for the better part of two decades before we agreed, roughly, on what it had taken from us — the frayed attention, the lost sleep, the pull we have ended up calling addiction. By the time we named the cost, the thing was load-bearing in every life, and it was far too late to set it down. That is the shape of this danger: not a catastrophe we could see coming and brace against, but a pattern we have already run once and are now running again on the scale of the self. The technology embeds before the alarm sounds. So the question is not whether we would be afraid if we could see this plainly. It is whether we will see it plainly while there is still something left to choose.