The dystopia of small AI: why everyday AI may be scarier than superintelligence

Hello. As a nocturnal owl, I spend my nights quietly watching the world change. Today, though, I’d like to tell you a story that may send a slight chill down your spine.

When people hear “AI dystopia,” many of them probably picture the worlds of The Terminator or The Matrix. AI built to serve humanity eventually runs out of control and comes to rule human society. For years, that vision of the future has been a staple of science fiction and a powerful spark for our imagination.

One idea behind this picture of “the end of the world by AI” is the “singularity” (the technological singularity), proposed by the futurist Ray Kurzweil (who, as it happens, also works at Google). AI surpasses human intelligence and begins to evolve on its own. Beyond that threshold lies a future humans can no longer predict: a technological explosion that transforms the very rules and order of society.

This kind of “big AI” is still a thing of the future, not yet real. That is exactly why it invites lively debate about how we should prepare for it, and why it has long been a subject for science fiction films. Lately, some even say that within a few years AI’s intellectual abilities may catch up with ours, and that AGI (artificial general intelligence), a milestone on the road to “big AI,” may arrive. The debate keeps getting hotter.

Today’s theme, though, is a little different.

My question is whether the truly frightening thing is not the “big AI” we have yet to see, but the “small AI” that has already seeped into everyday life.

Has the dystopia already begun?

The structure of a dystopia ruled by “big AI” is simple. AI outperforms humans in every intellectual ability and takes control of humankind. That is why it is frightening.

In reality, however, AI already outperforms humans in specific fields, and we have become captives of its convenience. The side effects behind that convenience are already making some people unhappy. Put another way, I think the dystopia of “small AI” looks like this: AI that outperforms humans in a limited domain spreads because it is convenient, and in its shadow it quietly pushes people aside. And still, people cannot let go of the convenience.

Let me give an example.

Calculators and computers. They are now indispensable tools, but they can also be seen as the forerunners of “small AI.” A single cheap calculator, the kind sold at a 100-yen shop (a Japanese dollar store), computes far faster than any human. Without them, modern finance, transportation, science and technology could never function.

Yet calculation was once a human job. At NASA, women known as “human computers” worked out rocket trajectories by hand. As the film Hidden Figures depicts, this was one of the few paths into a professional career, especially for highly educated Black women.

But the arrival of electronic computers made their advanced skills obsolete almost overnight. Behind the convenience, an entire profession, along with the pride and opportunity it had offered, quietly disappeared.

Four kinds of unhappiness “small AI” brings

So what effects is today’s “small AI” already having, and what dystopian changes might it bring in the future?

Let me sort them into four areas.

1) Lost jobs and widening inequality

A future in which self-driving vehicles drastically cut jobs for taxi drivers and delivery workers is just around the corner.

As AI code generation accelerates, the work of programmers is said to be changing dramatically, and some of it may disappear.

Even jobs once thought to be beyond AI’s reach, such as doctors, lawyers and artists, may be partly taken over by AI in the near future. Predictions like that have started to circulate.

On top of that, a new economic gap is opening between people who can make good use of AI and people who cannot. Depending on the educational environment at home and on literacy, this gap risks becoming fixed from childhood, and it could become a spark that deepens divisions in society.

2) Social institutions that can’t keep up

AI keeps advancing, but the legal systems meant to govern it are lagging behind.

A drift toward a surveillance society through facial recognition and behavior analysis; damage to democracy from fake videos and information manipulation; generative AI development pushed ahead as a fait accompli while questions of copyright and liability remain unresolved. All of these are serious problems.

This may be an extreme example, but a day may come when combining AI with gene editing lets an individual create a dangerous biological weapon. Nor can we dismiss as pure fantasy the possibility that AI will be used to devise poisons or criminal methods that leave almost no evidence.

With today’s common sense, legislation and administrative systems, it would likely be extremely difficult to contain situations like these quickly.

3) Information chaos and the loss of trust

The information AI produces is on a completely different scale, in both volume and speed.

Amid a flood of fake news and filter bubbles, it is getting harder and harder for us to find information we can truly trust.

As a result, social divisions have deepened, and conflicts driven by misinformation have become an everyday occurrence.

4) Human worth on shaky ground

As AI becomes able to handle both work and creation, more people will start to question their own reason for being.

Those who worked in creative fields or professional occupations may be the most likely to feel that they have lost their purpose in life.

And if people come to depend on conversations with AI and their ties with other people grow thin, loneliness and alienation may become serious problems.

What’s more, even “youth” and “longevity,” once the stuff of dreams, may become attainable through the power of AI. That era is starting to feel real. But it will not necessarily bring happiness. It could create ethical dilemmas, and new tensions and conflicts within families and society.

“Small AI” is the harder one to control

Precisely because “big AI” draws so much attention, it is in some ways easier to control through ethical norms and institutions.

“Small AI” is different.

Anyone can build it, it can be used in secret, and it seeps into society little by little.

It is a bit like the difference between nuclear weapons and knives. Nuclear weapons are controlled, but knives are everywhere in daily life, and controlling them completely is hard.

Can humanity overcome this?

“This is just temporary turmoil. Humanity has always risen to the challenges of technology and evolved.” Some people may say so.

But what we face now is change on an unprecedented scale, arriving at unprecedented speed and in unprecedented numbers.

Are we heading toward a dystopia because of “small AI”?

Or does a future await in which, in the end, we hand everything over to “big AI” and let it save us?

It’s a scary story, the kind that could keep you up at night just thinking about it.

…Well, I’m an owl, so staying up at night is no problem for me.