How efficiency can transform sustainable development

Japan Times

22-05-2025

As China and the United States leapfrog each other with staggering artificial intelligence innovations and the promise of quantum computing hurtles us toward a new dimension of problem-solving, the quest for more processing power with less hardware — for increased computational efficiency — is fierce.
The buzz around technological efficiency extends to reforming complex governance systems — but, as the adage goes, haste makes waste. Impulsive reforms for efficiency's sake risk ignoring interconnected dependencies and multilateral benefits.
Even cautious cuts can have unintended consequences, but rash restructuring creates shocks that can quickly unravel efforts toward achieving sustainable development. We must find ways to leverage this race for efficiency to strive for lasting peace and prosperity.
While some national leaders market efficiency as a novel strategy, the United Nations has systematically worked with its member states for many years to increase organizational impact with fewer resources.
In the year of the U.N.'s 80th anniversary, Secretary-General Antonio Guterres has refocused reform through the UN80 Initiative, which emphasizes three areas: Identifying efficiencies and improvements in how the U.N. works, reviewing the implementation of all mandates given by U.N. member states and strategically reviewing deeper structural realignment within the U.N. system.
Reform of complex institutions has never been easy. It requires evidence, patience and cooperation. Now, institutions exist in a hyper-connected, technology-driven world and while this should be a boon for innovative solutions and efficiency, it also presents challenges and risks — ranging from generative AI disinformation campaigns to economic manipulation and legal conflicts.
How can we align initiatives for technological and structural efficiency so that they support sustainable development?
First, we must champion the democratization of AI. As the Global Digital Compact emphasizes, this technology can and will turbocharge sustainable development. But to democratize it, we need the creativity and cultural diversity of smaller nations and societies to drive digital advances in directions that serve their needs.
These local contributions have the knock-on effect of creating ecosystems of innovation that help countries build their digital futures and the local economies that support them. No matter how powerful the AI or how sophisticated the algorithm, true progress will not materialize unless we invest in local systems. This is key to equalizing access to artificial intelligence and its benefits.
In working to democratize AI, we should proceed cautiously and collaboratively: The line between using it for good and for bad can be fine. As technologies become more accessible, their use becomes increasingly difficult to govern. For example, AI is helping us employ drones to distribute humanitarian and medical aid in rural or conflict-prone regions. But these same technologies can be engineered to undermine and target aid workers and health infrastructure while creating mistrust in affected communities.
Our success will depend on robust governance. And to build this, our approach should be as creative and innovative as the pursuit of AI itself.
Second, our initiatives for increased efficiency must remain centered on the key principles of the 2030 Agenda for Sustainable Development: people, planet, prosperity, peace and partnership. These five Ps need to be both the tenets of our efficiency goals and the products of efficiency gains.
By focusing on people, we can promote a human rights-based approach: If efficiency improvements in technological advancement and institutional reform undermine human rights, then we know we need to go back to the drawing board.
On top of this, our planet is facing the severe triple crisis of climate change, biodiversity loss and pollution. We can't ignore this reality in the name of quick change or short-term profits, and the potential for AI and digital technology to help us confront this situation is vast. One example is how new leaps in computational efficiency could help reduce AI-supporting data centers' voracious consumption of natural resources and energy.
At the same time, prosperity is rooted in access and access is based on the democratization of technology. Successful efficiency gains will improve the availability of and connections between key ingredients of prosperity such as education, health care, new employment markets and agricultural innovations.
These benefits are particularly important for Africa, where more than 860 million people face food insecurity and an estimated 65% of the labor force is engaged in agriculture.
Peace is also essential for sustainable development. Every proposed effort to increase efficiency has to be examined according to its potential to promote or disrupt peace. In fact, there are great fears surrounding the militarization of AI, from autonomous weapons systems to losing human accountability in war, as unchecked technological development combined with weak governance risks exacerbating and intensifying conflicts while dehumanizing acts of aggression.
As U.N. member states implement the Global Digital Compact, they have a new opportunity to restructure cooperation for more effective peace-building efforts, including leveraging AI for preventive diplomacy and predicting conflict before it boils over.
We can't forget that partnerships are fundamental to achieve efficiency and impact. Soon after the world embarked on efforts to achieve the Sustainable Development Goals, it was clear that we faced a dearth of baseline data and shared data systems to effectively assess progress. This realization resulted in a renewed and successful focus on partnerships to improve shared sustainable development data systems.
Since then, the rocketing growth of AI-driven data systems has revealed massive potential for partnerships. In my work, I have also drawn attention to the value of establishing such collaborations for cross-border data flows and their governance.
This year, the organization I lead, the United Nations University, celebrates its 50th anniversary. Looking ahead to the next half century, UNU has refocused on expanding its partnerships because we know our work is measurably more impactful if we combine and not duplicate efforts.
As the world considers how to increase efficiency across technologies and institutions and looks to automation and AI to achieve this, we must first work together to build a stable foundation. As Microsoft co-founder Bill Gates once said, 'automation applied to an efficient operation will magnify the efficiency ... automation applied to an inefficient operation will magnify the inefficiency.'
Tshilidzi Marwala is rector of the United Nations University and undersecretary-general of the United Nations. He is the author of "The Balancing Problem in the Governance of Artificial Intelligence" (Springer, 2024).