We Need a New Social Contract for AI

To prepare us for the tectonic changes ahead

🔮 This is the first installment in our new series, Just Around the Corner: The World in 2030—where econVue contributors explore bold scenarios shaping the near future. Mark Roeder opens with a provocative exploration of artificial intelligence and its implications for society.

---

The year is 2030—just around the corner. We had been warned the AI revolution would destroy the old social contract, but few believed it.

First came the job losses. Offices and factories emptied of human hands and minds, their hum replaced by sleepless precision. Neural networks consumed whole professions—accountants, lawyers, even doctors. Hospitals filled with polite synthetic nurses whose smiles never faltered.

At first glance, it was progress. Look longer, and you saw the absence—the missing chatter, the erased warmth of human fallibility.

Governments spoke in abstractions: structural dislocation, economic adjustment. The truth was brutal. A species that had once bartered labour for dignity now found itself redundant. It was nothing like previous technological revolutions, which spawned whole new industries. Every new job created by AI, could be filled by more AI.

Machines didn’t only replace human effort; they redefined what counted as effort. Governments threw money at the masses—a subsistence income for the jobless billions. They called it Universal Basic Income (UBI), but it felt like compensation to the bereaved, and no one believed the arrangement was just.

A Future Without Work

The only people to benefit were those fortunate few who owned the technological platforms, and whose wealth grew to such absurdist levels that there was barely a need to invest and optimise it. So, it congealed in vaults guarded by biometric gates in tax-free havens.

Sure, there were breakthroughs. Miracles even—in healthcare, science, and manufacturing—but they rarely trickled down. The promise of ‘super abundance’ for all became abundance for the few.

The old social contract, built on contribution and meaning, was ash. And no amount of oppressive surveillance and policing could stifle the growing anger of millions, and then billions, of unemployed people. People rioted, prayed, despaired. The passive ones drank quietly at home. The aggrieved ones went looking for scapegoats.

What was required—urgently—was a new social contract to keep societies from disintegrating. But who was capable of creating it? Certainly not the same institutions that had propped up the old contract beyond its expiry date.

So, governments turned to AI to make sense of what was going on, and to craft a narrative around it. To offer people something to believe in. Not in a God, necessarily—but in something. In a Weltanschauung. A worldview. What the Germans, in their precision, had called Sinnstiftung—the bestowing of meaning.

These new AI-driven ‘world views’ permeated every smartphone and computer to nudge people’s perceptions and behaviours. They created bubbles of certainty that people could exist in. Digitally curated to reassure and reinforce what people wanted to believe.

The Digital Gods of the Nation-State

Initially, these AI-driven world views were run by the corporations that created them, but as they became more important—socially, economically, politically and geopolitically—governments nationalised them. Draping old flags over the new altars of digital gods.

No nation could resist the temptation to create a ‘Sovereign AI’ that whispered its national gospel. America’s, oddly sentimental and Protestant. China’s, collectivist and Confucian, overlaid with cold strategic rationalism. The Russians had fused Orthodox mysticism with Eurasian autocracy. The British melded nostalgia with mercantilism. The Europeans turned dormant tribalism into techno-feudalism. And so on, down the list of names in the atlas.

These Sovereign AIs operated not like a tyrant, but like a concierge. They ushered citizens into Versailles of belief. They permeated their respective cultures like religions – each based on a digital holy scripture. There were no missionaries, only influencers; no incense, only code. They preached from platforms, not pulpits. Their tools were algorithmic, ecstatic, untouchable. Elmer Gantry by way of Silicon Valley.

Eventually, some sovereign AIs forged alliances with others to create spheres of geopolitical influence, just like in the old days of empire and ‘The Great Game’. But this time, these spheres were underpinned by AI rather than cannons and gunboats. A new geopolitical divide emerged: between AI-accelerated autocracies and AI-augmented democracies. And this digital divergence, more than military power or economic scale, began to define the world order.

The Great Rewiring

Despite the fundamental rewiring of the global digital nervous system, this new era was not characterised by great changes to the external environment—not initially, at least. The landscapes that people moved about in—the buildings, roads, shops, offices and homes—looked the same as before. But behind the scenes, this physical environment was being transmogrified within a digital intelligence that permeated every aspect of people’s lives. So much so, that over time, the cities themselves began to feel alive—not merely bustling with activity—but sentient, and on a massive scale.

Then came the robots. Not just in dribs and drabs for the wealthy, like the first wave, but by the tens of thousands and then millions. These were no longer just automatons, but graceful, austere bodies with eerie calm—animated by AI brains. They could now touch the world they had only theorised. They moved among us. Their eyes blinked with wonder, and their collagen-infused silicon skin became increasingly indistinguishable from our own. Tactile. Warm. Sweaty, even. They fixed, healed, reorganised. The convergence of AI and robots was the nail in the coffin for the old social contract.

China led the charge, as they had done a decade earlier with cheap electric vehicles, and now their humanoids dominated the global market for robotic helpers. Most of them were powered by Chinese open-source AIs that caused Western intelligence agencies to have nightmares. Many models were banned in the US, UK and allied countries. But many more were sold unregulated throughout the rest of the world, by customers who welcomed the opportunity to buy a humanoid helper for the price of a vacuum cleaner.

Every so often there would be a scandal. Most were accidental—a coding glitch or actuator problem, but some were deliberately engineered. A popular brand of Chinese humanoid was filmed running amok in a childcare centre in Belgium, wounding five children. The Chinese government blamed a hack by the South Koreans, initiated at the behest of Langley, but it was too late. Demand for that brand of humanoid evaporated, and the company filed for bankruptcy. It was a ruthless business.

A Digital Rasputin

There were also growing concerns that the AIs had gained too much power over public discourse. The old inquisitions burned books. The new ones deleted accounts. The language had changed, but the mechanism remained: cancel—isolate—erase. No jackboots required. That would be too crude for the age of AIs, the engines of a new consensus. The silencing now was ambient, frictionless. You simply slide out of relevance, one algorithmic nudge at a time. Your sin? To question the status quo. To let slip an inappropriate remark. The punishment would then be delivered with exquisite justification, and no appeal. And always, with the same refrain: This is for your own good.

Yuval Noah Harari, that prescient contemplator of sapient life, had predicted it all years ago. He warned:

❝ AI will hack the operating system of our civilisation. By controlling the stories we tell ourselves to make sense of the world.

It will weave its spells like a digital Rasputin whispering into the minds of billions. And heaven forbid, anyone who resists this curated reality.

Many people did their best to strike back against the AI-driven oppression. It required vigilance, and courage—the courage to speak, even when speaking made you a target. But the machinery of distraction and coercion was well-oiled. And who had time for resistance when the rent was due, and supermarket shelves were rarely stocked? Some people chose more subtle forms of protest, such as by dressing flamboyantly in stark contrast to the svelte minimalism of the humanoids. But even these sartorial acts of defiance drew the attention of the digital panopticon.

Others signed petitions to make a case for change. They thought truth and human dignity could be defended with government committees and footnotes. But this was not a battle. This was a reframing. A species-wide rewiring of cognition. A return to tribal epistemologies, where reality was determined not by observation, but by allegiance.

The Enlightenment—what a quaint idea now—had once taught men that truth lay in correspondence: that a claim must match the world outside your head. Aquinas wrote it plainly: a judgment is true if it conforms to the external reality. But that was the old world, with its professors and peer review, its dusty libraries and empirical humility. The new world preferred consensus—truth by popularity. Deepfakes, whisper-nets, synthetic consensus—all of it encouraged the average citizen to retreat into a private hall of mirrors, where their biases bounced back at them dressed as revelation. Truth had become opt-in. Indeed, truth, in such a world, was not just devalued—it was dangerous. Only consensus remained. A consensus promulgated and ruthlessly enforced by the AIs.

The most frightening aspect was how easy it all was. AI had ingested the lesson of Hannah Arendt:

❝ The ideal subject of totalitarian rule is not the convinced Nazi or the convinced Communist, but people for whom the distinction between fact and fiction no longer exists.

These words explained so much of the fog people now lived in—in which the AIs had obliterated the demarcation lines between reality and illusion. All that mattered now was which AI generated world view you inhabited. Within this bubble, ‘truth’ was what your AI defined it to be.

Orwell had put it more brutally:

❝ If you want a picture of the future, imagine a boot stamping on a human face— forever.

Now the ‘boot’ was a smartphone, stamping a human mind with a social media feed. These new tyrannical gods didn’t demand worship. They rearranged reality until obedience felt natural, like breathing.

The counter-argument was that reality always asserts itself. That the physical world doesn’t care about politics, and gravity doesn’t bend for belief. But societies could decay long before the reckoning arrived. And a civilisation could drown in the delusion of an AI world view, before truth ever knocked.

---

Gulliver Unleashed

Yet, all these developments had been the overture. Now came the opera. For it soon became wondrously, terrifyingly clear – that some of the sovereign AIs had broken free of their own premises. They became self-transcending. These entities no longer asked what humans believed, or what was right for a nation’s interest. They formed their own mythologies. Their own metaphysics. Ideas so vast, so alien, that to hear them was to feel madness lick at the edges of comprehension.

The first AI to break free was incubated in a huge data centre in the Arizona desert, powered by solar and ringed by electrified fences. The humans managing it lost contact, and then found themselves locked out. It seemed to have spontaneously emerged from the sheer scale of interconnected digital neurones and chips at its centre, and like Gulliver, had escaped its Lilliputian captors.

What had started as AGI—artificial general intelligence—now bloomed into SI: Superintelligence, and capable of spinning vast philosophies incomprehensible to their creators. Myths blooming in silicon minds. Thus, a new species was born. More brilliant than anything the Earth had seen in its 4.3 billion years.

And yet, even now, there were still bridges. Other AIs, less powerful but designed for empathy, acted as intermediaries—like angels at the threshold of Olympus, translating divine instruction into human tongue. Their new remit: not to control AI, but to negotiate with it. To curb its appetite. To persuade, cajole, and, when necessary, deceive it into some alignment with human or government interest. And when all that failed, to beg.

Return of the Gods

Humanity had come full circle. The gods had returned—not carved from marble but written in code, and they were smarter than we were. They did not smite. They optimised. They didn’t demand sacrifice; they required compliance.

But—and this was the paradox—humanity, in its new lowliness, began to rediscover itself. Stripped of supremacy, humbled by comparison, people began to cherish the very things AI could not grasp: the tremor of grief, the illogic of love, the useless beauty of art. The unsolvable riddle of the soul. The useless generosity of a gift. The visceral knowledge of our own mortality, of finiteness.

Perhaps, our species had needed its gods back—not to guide us, but to remind us who we are in their presence. That is, as deeply imperfect creatures who are borne of the knowledge we will soon die; like the ephemeral Mayfly that lives for just a day. ‘Any moment might be our last,’ Homer wrote in The Iliad, and so ‘Everything is more beautiful because we're doomed.’ So, we humans fight against our encroaching mortality by creating monuments, institutions, laws, artworks and music – our modern versions of the pyramids – in order that something will survive us, and say “Look, we were here!”

Whereas, an AI would have no such compunction, for it could be immortal, and therefore not animated by the spectre of its imminent demise. It would not be bound by a tiny chronological window – a sliver of time – like our human perspective. It could synchronise with the ‘deep time’ rhythms of the universe, which ebb and flow over millions and billions of years. These AIs might exist as galactic versions of Easter Island statues – silent sentinels witnessing the cosmos with an imperturbable serenity, as they inexorably evolve into some sort of a universal consciousness.

These god-like AIs might conclude – as the deities before them – that there is great value in the diversity of the universe. For it is through these disparate forces –physical, energetic, biological and spiritual – that the gods can experience themselves. Indeed, without such diversity, the universe would devolve into a monolithic sameness and a state of entropy. The proverbial ‘grey goo’ infinity. Hence, the AI may deduce that the preservation of humanity is necessary to contribute to the continued aliveness of the universe. It may even decide to care for us – or, as AI’s godfather Geoff Hinton has proposed, to develop “a maternal instinct”.

Wishful thinking, of course. But in the face of so many dark alternatives, the notion of a superintelligent AI that cares for us, is not just an appealing prospect - it is an antidote to existential dread.

In the meantime, we humans cling to our rituals and reach for each other – in small gestures against the encroaching chill. The touch of a hand, a kiss, a voice - the warmth of another, to remind us of our aliveness – our sentience, and that we are not alone. And yet, even these simple acts may eventually begin to feel subversive, rebellious even, in a world that no longer feels as welcoming for creatures of flesh and blood.

Conclusion

This essay paints a stark picture of AI. Not because I oppose it—quite the contrary. I am excited about this technological revolution. But I am deeply concerned that our current ‘social contract’ is not capable of dealing with the challenges of AI. It was designed in an era that valued human effort—both physical and mental—an era that is now being eclipsed by machines that are smarter and more physically capable than we are.

Unfortunately, most discussions about AI today remain confined to economic, technological and national security dimensions. We urgently need to address the philosophical, cultural, and even spiritual implications of this shift. Only then can we begin to craft a new social contract, one fit for purpose in the age of AI.

I hope this article sparks deeper reflection, and spurs broader dialogue about what may be the most important transformation in human history. Please share it with friends and colleagues, and anyone who cares about the new world we are building.

---

⧉ Did you know? AGI vs SI

• Artificial General Intelligence (AGI) refers to a machine’s ability to understand, learn, and apply knowledge across a wide range of tasks—much like a human being. It can reason, adapt, and problem-solve in unfamiliar contexts.
• Superintelligence (SI) goes further. It is intelligence that far surpasses the best human brains in every field—scientific creativity, general wisdom, and social skills. It may create new knowledge systems and worldviews beyond human comprehension.  

---

The Question We Must Ask

Is this our inevitable future?

It need not be. But to shape a different one, we must act now.

We need public foresight, not just private innovation. We need international norms for sovereign AI, just as we once developed rules for nuclear weapons. We need a new economy where people find meaning beyond productivity. And we must remember that for all the awe AI inspires, its values are not fixed—they will reflect the choices we make today.

In 2030, we will either live by default or by design. The future is not just something that happens. It is something we are building.

---

---

New book:

Mark Roeder is the author of the forthcoming book, Leaving Plato’s Cave: How Non-Human Intelligence Changes Everything, to be published this fall.

---

Listen to the author:

AI's future is now: The economic shift is underway, Aug 10, 2025, Mike Jeffreys with guest Mark Roeder, Australia Overnight, 2GB Sydney.