New Great Divergence in the Age of Artificial Intelligence: Risk of a New Global Divide
The impact of artificial intelligence is not uniform across all countries. Making the gap more uneven is exceptionally likely if the governments don't act urgently in a coordinated manner.
Deep Inequalities in world inequality regarding income, skills, infrastructure, and institutions are the focus of the next great divergence warning. AI technologies have no inclusive policies to help the world. They leave further behind all countries that are not advanced.
Shifts in the AI and development debate are overdue. Isolation of technological advancements means that the focus is solely on productivity or the fear of automation. They mostly ignore the damage that inequality inflicts.
A more developed digital structure, a more advanced data ecosystem, and a developed research capacity will provide the overwhelming bulk of the value to countries. On the contrary, they will only become less developed, and historical patterns will further confirm this negative impact.
Technology is always diffusing worldwide, but it does not do so in an even manner. This leads to long-term divergence in country productivity and income.
AI can increase human productivity by automating many routine tasks and even enhancing decision-making. But it is worth contemplating the fact that the productivity crunch is only focused on a few specific sectors. Emerging economies run the risk of being squeezed further.
More recent studies are confirming this phenomenon. High-income countries with strong digital skillsets and R&D ecosystems are seeing the greatest benefits from the adoption of AI. Conversely, the developing nations of the world have been unable to turn that same access into productivity gains.
The greatest impact will be on human capital. AI-based tools have the ability to enhance education and health diagnostics and improve agricultural productivity through personalized learning, early disease detection, and predictive insights for farmers.
But this access is often extremely unequal, with gaps in digital literacy and device ownership affecting women in particular in many areas of Asia. Rural populations, especially, lack the basic digital skills required to use even the simplest productivity tools. Countries that report lower levels of digital literacy and weaker education systems tend to experience slower AI adoption and weaker economic benefits of AI.
The scope of governance of the country will further delineate the outcomes. Countries that possess robust regulatory frameworks can utilize AI to enhance the efficiency of the public services they provide, manage risks, and provide further transparency.
Others are likely to be reliant on regulated, un-auditable, and unreliable (from the governance of the country perspective) imported technologies. This scenario creates a chronic dependency and undermines the flexibility of policy-making.
AI governance research suggests that the lack of robust technical capacities within a state will lead the state to rely on unaccountable, opaque, and private systems that are susceptible to the widening of the gap of inequitable access to services.
The added layer of environmental and infrastructural constraints is self-explanatory. In order to train large AI models, large quantities of energy and water are required. By 2030, data centers are expected to use more than twice the water and nearly triple their electricity consumption.
Countries with low digital literacy and weak education systems are slower to adopt new AI technologies. Countries with weak energy facilities and limited resources will face challenges in developing domestic AI ecosystems. This increases dependency upon foreign AI ecosystems and platforms. The main point is that- technology by itself does not lead to progress. It is policies that do so.
Countries that invest in digital infrastructure will reap the benefits. Countries that do not will have worse issues. The evidence clearly illustrates the need for action. Governments need to see AI as a development policy in itself, not just technology.
First, they need to build basic infrastructure. People need to have access to electricity and the internet, and devices that are not costly. This is necessary for inclusive adoption.
Second, they need to invest in education. Digital literacy, data skills, and AI education need to be part of the education system.
Third, they need to have better regulations. Clear and transparent systems can avoid reliance on systems that are not. Fourth, they need to promote cooperation amongst each other. Standards that are shared and cross-border data governance can promote cooperation.
The possibilities of Artificial Intelligence (AI) are both very promising and perilous. While AI has the potential to boost inclusive growth, improve public service delivery, and accelerate progress, it can also establish and perpetuate a new era of divergence.
To avoid the negative potential of AI, policymakers need to implement crucial and decisive actions right away. Governments need to act constructively, plan their investments wisely, and manage/oversee their actions responsively.
If they do not implement these structures, the next technological revolution will most likely add to the existing global inequalities. Conversely, if they manage to implement these structures, they could allow the use of AI as a dependable resource to bring convergence and avoid divergence.
Feroz Ahmed is a Research Assistant at Lamar University, U.S.A. His work focuses on management information systems and digital governance, with work spanning AI, data-driven policy, and digital transformation.
