Can the World Meet AI’s Resource Demands?

This Sky News Arabia analysis explores whether the world can keep up with the explosive resource demands of generative artificial intelligence. As AI becomes a central driver of innovation, it behaves like a “hungry giant,” consuming more data, chips, electricity, water, land, and skilled human labor. Oxford-certified AI expert Hilda Maalouf Melki frames the key question: how can societies satisfy AI’s appetite for natural resources without harming the planet?

Electricity: The Core Bottleneck

A major focus of the article is electricity. AI systems depend on vast networks of data centers filled with servers, storage devices, and networking equipment. In many countries, demand from these centers already exceeds available grid capacity.

Projections cited in the article suggest that data centers in the United States alone could consume about 8% of national electricity by 2030, roughly triple their share in 2022. Globally, data center energy use is expected to reach levels comparable to the total electricity consumption of large countries like India.

Large technology companies are aware that energy may become the most serious constraint in AI’s supply chain. They are rushing to secure long-term electricity contracts and explore new sources of power.

Water and Cooling: Hidden Environmental Costs

Beyond electricity, AI infrastructure uses enormous quantities of water for cooling. Modern data centers generate intense heat and must be cooled efficiently to protect equipment and maintain performance.

The article notes estimates that data centers consume more than a billion liters of water per day worldwide—enough to meet the daily needs of millions of people. Much of this water is potable, raising concerns about competition with human and agricultural needs, particularly in water-stressed regions.

Even everyday AI usage has a footprint: one study estimates that a typical ChatGPT session of 10–50 questions can consume the equivalent of a small bottle of drinking water when cooling is accounted for.

Land and Real Estate: A New Spatial Race

AI infrastructure also reshapes the geography of land use. The number of data centers worldwide has roughly doubled over the past decade, and thousands more are expected as AI capacity expands.

Data centers require large, strategically located plots of land near reliable power, fiber networks, and sometimes cooler climates. This demand can drive up property prices, spark competition for prime sites, and alter local development patterns. The article warns of potential “real estate wars” as major tech players seek the best locations.

Hilda Maalouf Melki notes that a smarter approach may be to build centers in deserts or mountainous regions less suitable for agriculture or housing, and to retrofit existing industrial buildings rather than consuming new urban land.

Chips: The Battle for GPUs

The article highlights how advanced chips—particularly GPUs—have become another choke point. These processors are essential for training and running the latest AI models, but global supply remains constrained.

High-end chips can cost tens of thousands of dollars each, and AI systems often require hundreds or thousands of them. Even major suppliers struggle to keep pace with demand, creating bottlenecks that slow AI projects and intensify geopolitical competition around semiconductor supply chains.

This scarcity underscores that AI is not just a software revolution; it depends on physical components with their own production limits, security risks, and environmental costs.

Jobs, Talent, and New Labor Pressures

Public debate often focuses on how AI might replace human workers. The article acknowledges those concerns, citing estimates that AI could affect up to 40% of jobs worldwide.

Yet Hilda Maalouf Melki stresses that AI is also generating new jobs in fields like data science, software engineering, chip design, legal compliance, and digital policy. The problem, she argues, is a mismatch between demand and supply: there are not enough qualified people to fill these roles.

This creates “talent bottlenecks,” especially in highly technical fields. Her proposed response is not to slow AI, but to invest heavily in education, training, and upskilling—particularly in regions that risk being left behind. She highlights the importance of making advanced technical education more accessible so that more people can participate in and shape the AI economy.

Focusing on AI’s Positive Potential

Despite the repeated emphasis on resource strain, the article closes on an optimistic note. Hilda Maalouf Melki argues that AI, if governed wisely, can deliver major benefits: faster medical innovation, improved education, more efficient industries, smarter sustainability solutions, and better quality of life.

For her, the central task is to stop viewing AI only as a threat and instead treat it as a powerful tool that must be aligned with human values, environmental limits, and long-term public interests.

Rather than asking whether the world can afford AI, she suggests a different question: how can we adapt our energy systems, water management, education, and planning so that this “hungry giant” becomes a driver of sustainable progress rather than a source of crisis?