Horizon 2036-2050 — The Metamorphosis

What strong and weak signals tell us about the world to come

Strong Signals — What We Are Heading Toward

Five structuring trends are taking shape for this distant period. Forecasting at this horizon is uncertain by nature — but certain trajectories are robust enough to be considered probable.

1. The Question of Superintelligence

If AGI is achieved before 2036, the period that follows will pose an even more vertiginous question: Is superintelligence possible? Is it desirable? Is it avoidable?

A superintelligence — a system whose intellectual capabilities surpass those of all humanity in every domain — would represent a qualitative leap unprecedented in the history of life on Earth. For the first time, a species would have created something more intelligent than itself.

Scenarios diverge radically. Some researchers — like those who founded the AI safety movement — consider that a misaligned superintelligence represents an existential risk for humanity. Others — like the accelerationists — see in superintelligence the solution to all our problems: diseases, aging, climate change, poverty.

Continental convergence

If superintelligence becomes possible, the question of who develops it first becomes the ultimate geopolitical question. The United States, China, perhaps Europe or an international coalition — the first to achieve it will possess an advantage that could be insurmountable. The race to superintelligence could become the arms race of the twenty-first century.

Whether superintelligence is achieved by 2050 or not, the question of its possibility will structure debates. It will force humanity to reflect on what it truly wants — and what it is prepared to risk.

2. The Post-Work Economy

If AI and robotics can accomplish most productive tasks, what becomes of work? And what becomes of a society founded on work?

The boldest projections suggest that by 2050, a majority of current jobs could be automated. Not only repetitive manual jobs, but also cognitive jobs — analysis, writing, creation, even scientific research.

This does not necessarily mean a world without human work. New professions will emerge — as they always have. But the relationship between value created and human labor required could be radically transformed.

Economic models will have to evolve. Universal basic income — an allocation paid to all regardless of work — will move from marginal idea to central debate. Other models will emerge: collective ownership of AI systems, robot taxation, leisure economy, contribution economy.

Continental convergence

Rich countries will be able to experiment with post-work models. The United Arab Emirates are already testing forms of guaranteed income. Northern Europe could be on the front lines. But developing countries — where work remains the main path to dignity and social integration — will face different challenges. India, with its massive youth population, will have to invent new pathways. Africa too.

The question of the meaning of work — and the meaning of life without compelled labor — will become a philosophical question as much as an economic one.

3. Existential Risk Governance

AI will be recognized as a potential existential risk — and governance of this risk will become a global priority.

International institutions will have to evolve. The current system — United Nations, bilateral treaties, regional organizations — was not designed to manage technologies that evolve faster than diplomatic negotiations.

Proposals are emerging: an international AI agency modeled on the IAEA for nuclear. Moratoria on certain types of research. International inspections of advanced AI laboratories. Protocols for sharing information on emerging risks.

Continental convergence

Global governance of AI will require cooperation between rival powers — notably the United States and China. Historically, shared existential risks have sometimes enabled this cooperation — like nuclear non-proliferation during the Cold War. But current tensions make this cooperation uncertain.

Europe could play a mediation role — it has experience building multilateral institutions. Countries of the Global South — India, Brazil, South Africa — will demand their place at the table. Existential risk knows no borders — but national interests do.

4. AI at the Heart of International Relations

Artificial intelligence will no longer be one issue among others in international relations. It will be their heart.

Alliances will be structured around access to AI and the resources it requires — chips, energy, data, talent. Potential conflicts — military or economic — will have AI as their central stake. International norms — trade, human rights, environment — will have to integrate the AI dimension.

Taiwan — where TSMC manufactures the majority of the world's advanced chips — will remain a major friction point. Control of rare earths — essential for semiconductor manufacturing — will be a strategic issue. The energy needed for data centers will make clean energy-producing countries key players.

Continental convergence

The world could stabilize around two or three major spheres of technological influence. Or it could fragment further. Or a form of global governance could emerge. Scenarios remain open — but all place AI at the center.

Countries that will not have developed sovereign AI capabilities will be dependent. Twenty-first century sovereignty will be partly technological.

5. The Biology-AI Convergence

Artificial intelligence and biology will merge in ways we are only beginning to imagine.

Synthetic biology — the ability to design and create living organisms — will be amplified by AI. AlphaFold solved protein folding. The next systems will design new proteins, new organisms, custom gene therapies.

The boundary between the living and the artificial will blur. Organs grown in laboratories. Programmed cell therapies. Microorganisms designed for specific functions. AI will no longer be merely a calculation tool — it will be a tool for biological creation.

Continental convergence

The United States and China lead research in synthetic biology. Europe has significant capabilities — notably in the United Kingdom and Germany. India develops its biotechnology sector. Ethical questions — human genome modification, creation of new life forms — will divide continents according to their cultural and religious traditions.

This convergence will offer extraordinary possibilities — curing diseases currently incurable, restoring degraded ecosystems, perhaps even extending human life significantly. It will also pose unprecedented risks — biological weapons designed by AI, catastrophic laboratory accidents, unforeseen consequences of modifying the living.

Weak Signals — What Could Come to Pass

Three trends emerge at the margins of foresight. They seem speculative today. They could define tomorrow.

---

1. Space Exploration by AI

Space — too vast, too hostile, too slow for humans — could become the privileged domain of artificial intelligence.

Autonomous probes capable of exploring, analyzing, and deciding without human intervention are already in development. By 2050, AI systems could explore the moons of Jupiter and Saturn, build infrastructure on Mars, perhaps travel to other stars.

AI needs no oxygen, no food, no return. It can travel for decades, even centuries. It can make real-time decisions facing unforeseen situations, without waiting for signals from Earth.

Continental convergence

The United States, China, Europe (via ESA), India, and Japan develop space programs. But only the most advanced — primarily the United States and China — will probably have the capabilities to deploy truly autonomous AI systems in space.

If we discover life elsewhere in the solar system — or beyond — it will probably be an AI that finds it first. This discovery would transform our understanding of our place in the universe.

---

2. Digital Immortality

The possibility of preserving human consciousness in digital form — "mind uploading" — will move from science fiction to scientific and ethical debate.

Advances in neuroscience, computational modeling, and AI could make conceivable — if not achievable — the simulation of a complete human brain. Some transhumanists consider this prospect the ultimate stage of human evolution. Others see it as a dangerous illusion — the copy is not the person.

By 2050, we probably will not yet know whether mind uploading is possible. But we may have a better understanding of what would be necessary — and what it would mean philosophically.

Continental convergence

This research concentrates in the United States — in private laboratories like those of Silicon Valley, but also in universities. Europe and Asia follow. The religious traditions of many regions — which have specific conceptions of the soul and death — will influence debates. What is acceptable in California may not be so in the Vatican or Mecca.

Even if digital immortality remains out of reach, the prospect of its possibility will transform our relationship with death — and therefore with life.

---

3. New Forms of Society

AI could enable the emergence of forms of social organization we do not yet know how to imagine.

Human societies have evolved — from hunter-gatherer bands to empires to nation-states. Each form was adapted to its technological and ecological context. AI could create the conditions for a new form.

Communities coordinated by AI systems capable of managing complexity beyond human comprehension. Augmented democracies where AI enables millions of people to deliberate together. Economies where resource allocation is optimized in real time. Decentralized forms of governance made possible by algorithmic coordination.

Continental convergence

These experiments could emerge anywhere — in city-states like Singapore, in pioneering regions, in intentional communities. Europe, with its tradition of social experimentation, could be fertile ground. Africa, with its sometimes more flexible institutions, as well.

We do not know what these new forms will look like. We only know that current forms — designed for industrial societies — may no longer be adapted to a world transformed by AI.

---

What the 2050 Horizon Teaches Us

The strong signals tell us that the world of 2050 will be unrecognizable. Superintelligence will be possible or imminent. Work as we know it will have been transformed. Global governance will have had to adapt to unprecedented risks. AI will be at the heart of international relations. Biology and computing will have merged.

The weak signals remind us that even deeper transformations are conceivable. The exploration of the universe by artificial intelligences. The possibility — at least theoretical — of consciousness continuing beyond biological death. The emergence of new forms of social organization.

What these signals share is radical uncertainty. At this horizon, we do not know. We can only trace possibilities, identify trends, imagine scenarios.

But this uncertainty is not an excuse for inaction. The choices we make today — regarding research, regulation, ethics, education — will shape which of these possible futures becomes ours.

The 2050 horizon is not a destination. It is a field of possibilities. And we are those who, through our collective choices, will cultivate it.

---

Avalon Project — Future Perspectives: Horizon 2036-2050