TECH

AI is growing, but can the planet handle it? The hidden cost of the digital revolution

Artificial intelligence is emerging as a hero in the fight against climate change, but its accelerated growth hides an environmental cost that almost no one sees. Between explosive energy consumption, thirsty data centers, and a new avalanche of electronic waste, the technology that can save us also threatens to worsen the crisis it is trying to solve.

Artificial intelligence is now at the center of major solutions to the planet's challenges. It anticipates catastrophes, optimizes energy systems, creates more sustainable materials, and revolutionizes science. However, behind this promising image, an invisible physical structure is growing that requires gigantic volumes of energy, water, and natural resources. As AI expands, its own environmental footprint is beginning to raise a global alarm.

Training large AI models requires colossal amounts of electricity. To give you an idea, large-scale text generation systems have already consumed volumes of energy equivalent to the annual consumption of dozens of homes. Today, with daily use on a global scale, this consumption is even greater.

This growth has led tech giants to seek their own energy sources, including contracts with nuclear power plants, to ensure the stable operation of their data centers. The problem is structural: the architecture of current computers constantly transfers data between memory and processor, generating heat and wasting energy. With physical limits approaching, efficiency is no longer growing at the same rate as demand.

Water, minerals, and waste: the invisible chain of AI...The impact of artificial intelligence goes beyond the electricity bill. Data centers use enormous volumes of potable water to cool equipment that operates non-stop. On average, each kilowatt-hour consumes liters of clean water, a resource that is increasingly scarce in many regions of the world.

In addition, the manufacture of chips requires rare minerals, often extracted under controversial social and environmental conditions. At the end of the life cycle of this equipment, another problem arises: the explosion of electronic waste, which is difficult to recycle and highly polluting.

Artificial intelligence lives in a profound contradiction. While contributing to the energy transition, precision agriculture, climate monitoring, and disaster prevention, it also increases the pressure on the planet's own resources.

Studies indicate that AI can help advance many global sustainability goals, but it can also delay several of them if its growth goes unchecked. It's not about rejecting the technology, but about recognizing that its impact is ambiguous.

The solutions are not just in more efficient software. The real leap will come from hardware. Research is advancing in in-memory computing, which eliminates unnecessary energy shifts, in memristors that process and store simultaneously, in photonic chips that use light instead of electricity, and in analog systems inspired by the workings of the human brain.

These technologies promise to drastically reduce energy consumption and device heat generation.

Governance: the decisive factor for the future of AI...The sustainability of artificial intelligence depends not only on technical innovation, but also on public policies, transparency, and corporate responsibility. "Green algorithm" programs and environmental requirements are already beginning to emerge in some countries.

AI has the potential to profoundly transform the world. But this transformation will only be truly positive if the technology itself is designed not to become yet another threat to the planet's climate balance.

The growth of artificial intelligence (AI) represents a significant challenge to the planet's sustainability due to its high consumption of energy and water resources, especially in the data centers that support it. However, AI also offers potential for energy efficiency solutions.

The Hidden Cost of AI

-Massive Energy Consumption: Data centers, the heart of AI infrastructure, already consume about 1% to 2% of global electricity. Intensive AI use could increase this consumption dramatically; some estimates predict that data center electricity consumption could more than double by 2030. This increase in demand often relies on non-renewable energy sources, contributing to greenhouse gas emissions.

-Intensive Water Use: The cooling systems needed to prevent servers from overheating consume vast amounts of water. Large data centers can use millions of gallons per day, the equivalent of the consumption of a medium-sized city. It is estimated that 20 to 30 questions to a generative AI can consume half a liter of water.

-Resource Extraction and Electronic Waste: The production of AI hardware requires the extraction of valuable minerals, such as lithium and cobalt, which have significant environmental impacts. The rapid planned obsolescence of this equipment also exacerbates the problem of electronic waste.

The Potential of AI for Sustainability...Despite the costs, AI can also be a powerful tool in the search for climate solutions:

-Energy Optimization: AI can optimize energy consumption in buildings, transportation networks, and data center management, increasing operational efficiency and reducing waste.

-Climate Forecasting and Resource Management: AI models can help predict climate patterns, manage water resources more efficiently, and better integrate renewable energy sources into the electricity grid.

mundophone