Imagine that artificial intelligence delivers much of what its most vocal advocates promise: businesses become dramatically more productive, routine administrative tasks are accomplished in minutes rather than hours, and scientists discover promising new drugs and materials at unprecedented speeds. Energy networks become vastly easier to manage, and a small, regional enterprise gains analytical capabilities that previously required an entire corporate department. By familiar, traditional economic measures, there would be an immense amount to celebrate.
Yet, when we widen the picture, a series of critical questions immediately emerges. Are human beings actually becoming more capable of shaping their own lives? Are these profound technological gains reaching the communities and individuals who need them most? Are the natural systems that sustain life on Earth becoming healthier? And as machines grow increasingly adept at writing, analyzing, remembering, and recommending, are they genuinely strengthening human capability, or are they quietly encouraging us to exercise less of it?
These are not merely philosophical musings; they are fundamentally economic questions. They point directly toward what may well become the next frontier of human economic progression: an architecture that becomes far better at generating more of what truly matters, by recognizing that social, economic, ecological, and technological progress are deeply and inextricably intertwined.
One Economy, Four Connected Systems
Modern societies have grown accustomed to dividing reality into manageable, distinct administrative pieces. Economic ministries deal strictly with growth, gross domestic product, and market efficiency. Social ministries address health, education, and the persistent challenges of inequality. Environmental agencies focus narrowly on climate change, biodiversity, and pollution control. Meanwhile, technology policy concentrates heavily on innovation, intellectual property, and global competitiveness. While these divisions make administrative and bureaucratic sense, life itself is far less cooperative.
Consider something as physically grounded as a modern data center. Economically, it represents substantial capital investment, technological infrastructure, and productive capacity. Technologically, it provides the immense computing power required to train and run advanced artificial intelligence models. Socially, its ultimate significance depends entirely on what that computational power enables, which jobs it transforms, and who reaps the rewards. Ecologically, that exact same facility demands a massive, continuous supply of electricity, water, specialized equipment, rare earth minerals, and land. A single investment moves through four distinct systems at the exact same time.
The numbers make this convergence strikingly tangible. Current global energy projections suggest that the world’s data centers could consume around 950 terawatt-hours of electricity annually by the year 2030. That figure represents nearly twice their projected consumption for 2025—an amount of power slightly larger than the total annual electricity consumption of Japan today. While data centers will still account for roughly three percent of projected global electricity demand, the real issue is not that artificial intelligence will somehow consume all the world’s power. Rather, the challenge lies in the sheer speed of expansion and the intense, localized pressure that concentrated demand places on particular electrical grids.
This rapid technological acceleration is unfolding on a planet that is already under severe ecological strain. The Planetary Health Check report indicates that seven of the nine planetary boundaries have been pushed beyond their safe operating space. In everyday language, several of the complex natural systems that have historically provided relatively stable, predictable conditions for human societies are now being pushed beyond scientifically assessed limits.
Artificial intelligence could theoretically help alleviate these pressures. It can actively balance complex electricity networks, improve industrial efficiency, reduce supply chain waste, accelerate the discovery of sustainable materials, and monitor forests, oceans, and agricultural lands with unprecedented precision. Technology has the dual capacity to both increase ecological pressure and help reduce it; to widen access to economic opportunity while simultaneously concentrating economic power; and to relieve people of tedious, repetitive work while potentially removing the very experiences through which human expertise and mastery develop.
These effects are profoundly connected. That is why improving one isolated part of the system while transferring hidden costs to another closely resembles moving a lingering expense between columns of a household budget and declaring ourselves richer.
Where Does the Productivity Go?
Artificial intelligence makes this systemic evaluation urgently necessary because measurable productivity gains are already appearing in workplaces across the globe. A large-scale workplace experiment found that regular users of generative artificial intelligence spent about two fewer hours each week managing email and successfully reduced the amount of work they conducted outside of normal business hours. In practical terms, artificial intelligence created time.
The central question is what happens to that newly created time.
An hour saved through automation can easily become simply another hour of standard production. However, it can also become an hour dedicated to continuous learning, caregiving, creative exploration, better customer interaction, or simply time returned to an employee for rest and personal well-being. Similarly, higher productivity can translate into lower consumer prices, larger corporate profits, higher wages, greater investment in social infrastructure, or better-funded public services.
Which direction it ultimately goes depends heavily on economic systems and the human choices those systems incentivize. This directly links to another critical asset that traditional economic statistics capture very poorly: human agency—our fundamental capacity to understand, judge, choose, and act.
Artificial intelligence possesses the unique capacity to both strengthen human agency and encourage us to exercise it far less. Global Positioning System technology offers a familiar, everyday analogy: personal navigation has become wonderfully efficient, yet it is now entirely possible for individuals to travel repeatedly through a city without developing any meaningful understanding of its geography or layout.
Apply that exact same principle to core human activities like writing, remembering, and making complex decisions, and the stakes become significantly higher. In a randomized experiment, artificial intelligence substantially narrowed performance differences between people with varying educational backgrounds while the technology was actively available. However, when participants later worked without the assistance of artificial intelligence, those performance gains did not translate uniformly into independent capability. How people engaged with the technology mattered profoundly. Consequently, the core economic challenge is far more sophisticated than simply deciding whether humans should delegate tasks to machines.
Making Progress Work Harder for People and Planet
This is where the concept of ProSocial AI becomes vital. It is not about creating machines that are merely friendly, polite, or altruistic, but rather about developing and deploying artificial intelligence with a much wider, more holistic understanding of value. This means using technological capability to systematically improve outcomes across people, shared prosperity, and the planet, rather than optimizing a single narrow measure while ignoring detrimental effects elsewhere.
The potential applications can be thoroughly practical and grounded in reality. An agricultural system can help a farmer optimize crop yields while simultaneously reducing unnecessary water usage and chemical fertilizer runoff. Artificial intelligence can remove administrative paperwork from a nurse’s daily routine and return valuable time directly to patient care. A logistics network can simultaneously reduce corporate costs and environmental waste. A teacher can accurately identify where a specific student needs academic support while still ensuring that the student performs the mental work required to genuinely learn and retain knowledge.
The common thread running through all of these examples is better alignment.
This leads naturally toward Hybrid Intelligence, a deliberate, thoughtful combination of natural and artificial intelligences structured so that their respective strengths complement one another. Human beings contribute lived experience, core values, moral responsibility, social understanding, and complex judgment. Artificial systems contribute processing speed, massive scale, and the ability to process quantities of information that no single person could reasonably manage. Both systems operate within the natural environmental boundaries upon which every human economy ultimately depends.
The ultimate destination of this approach is Hybrid Flourishing—a condition in which technological progress actively expands the capacity of both people and the planet to thrive, giving every living being a fair chance to fulfill its inherent potential. This state grows from the inside out through human values, choices, and agency, while being nurtured from the outside in by public institutions, fair markets, and technologies that support those choices.
Society cannot simply abandon economic growth; billions of people across the globe still urgently need better housing, improved healthcare, quality education, reliable infrastructure, and fundamental economic security. Material progress remains indispensable. The fundamental task for modern society is to ask more of it.
For two centuries, economic development has become progressively better at producing more material goods, while steadily valuing humans and nature less. Injected directly into the current economic architecture, unguided artificial intelligence will simply accelerate that existing achievement. Yet there is a clear alternative: creating prosperity in ways that actively strengthen the people, societies, and living systems from which all true prosperity originates.
Society already knows how to make machines more productive. The defining opportunity of this generation is making progress more productive for life itself.