ARTIFICIAL INTELLIGENCE: THE MACHINE THAT CONSUMES THE PLANET

Artificial intelligence (AI) is presented as a major technological revolution, yet its functioning depends on an extensive industrial ecosystem. This article examines these aspects using available data, without ecological exaggerations, in order to provide a clear picture of the technological impact.

The Endless Hunger for Metals

AI relies on the massive extraction of rare metals, leaving a significant ecological footprint. It requires large quantities of copper, cobalt, lithium, nickel, and rare earth elements. These resources are extracted through mining operations that can affect the environment and local communities. Every new AI chip requires raw materials, leaving behind a measurable ecological footprint.

The Invisible Factories of the 21st Century

The data centers that support AI consume large amounts of energy and can generate high emissions when powered by fossil fuels. Behind every AI‑generated response operate enormous data centers where thousands of processors run continuously. These facilities have extremely high energy consumption, sometimes comparable to that of medium‑sized cities. As models become more complex, the demand for electricity increases, often sourced from fossil fuels.

Water, the Unseen Fuel

The data centers that support AI use large amounts of water to cool servers, which can strain local resources. Few users realize that generating a digital response also involves water consumption. Servers processing large volumes of data produce significant heat and rely on cooling systems to maintain safe operating temperatures. Many of these systems depend on water, creating pressure on local supplies, especially during droughts.

An Industry That Demands More and More

Generating AI responses consumes far more energy than a standard web search, and at the scale of billions of interactions, this consumption becomes significant. Search engines display existing information retrieved from web pages. An AI model generates new content by calculating probabilities and relationships across billions of parameters. This process requires more computational power than a typical search. When millions of users interact with such systems daily, the cumulative energy consumption becomes substantial.

The Obsession with Ever‑Larger Models

Companies are developing increasingly large AI models, amplifying energy and resource consumption across the industry. They compete not only for performance but also for size. Models grow larger, require more processors, and demand ever‑greater investments. This technological competition creates constant pressure to expand infrastructure, leading to rising consumption of energy and natural resources. Every actor in the industry tries not to fall behind, resulting in a continuous spiral of growth and consumption.

The Myth of the “Green Cloud”

The infrastructure that supports the “cloud” is, in reality, a heavy industry hidden from the user’s view. The tech sector often speaks about sustainability, yet the infrastructure behind the cloud is entirely material. Every data center requires thousands of tons of concrete and steel, kilometers of cabling, industrial cooling systems, transformers, generators, and dedicated electrical networks. There is nothing immaterial about this structure — it is heavy industry, placed out of sight.

The Unseen Cost

The user sees only the quick response, not the hidden industrial chain that consumes resources and energy to make AI function. The millions of fans running continuously are invisible, and carbon dioxide emissions remain unseen precisely because the gas is colorless. The experience feels smooth and simple because the infrastructure supporting it is completely hidden.

Technology Is Not Free

AI progress relies on a global industrial chain that consumes finite resources and leaves behind invisible environmental costs. AI is presented as a symbol of progress, yet it is supported by a very concrete infrastructure. Every digital response is part of a global industrial process: it begins with mineral extraction, continues in semiconductor factories, passes through large‑scale data centers, and finally reaches the user’s screen. In this process, finite resources are consumed — minerals, electricity, water, and land. Environmental costs are incurred long before the user receives the response, even if they remain invisible.

An Uncomfortable Question

AI provides real benefits, but using it for low‑value content raises ethical questions about resource consumption in a world with clear limits. AI can bring significant advantages in fields such as medicine, research, and education. At the same time, a substantial portion of its daily use produces low‑value content: videos, advertisements, or texts that quickly lose relevance. In an industry that consumes natural resources on a global scale, the discussion is no longer only about technological capabilities, but also about how we choose to use them. In the context of climate change, drought, and pressure on resources, every request sent to an AI model represents a decision about how we use the planet’s limited resources.

Our Summary

1. AI depends on the massive extraction of rare metals, leaving a significant ecological footprint.

2. The data centers that support AI consume large amounts of energy and water, generating high emissions when powered by fossil fuels.

3. Generating digital responses consumes far more energy than a standard search, and at billions of interactions, this consumption becomes immense.

4. Companies are building increasingly large AI models, amplifying resource consumption across the industry.

5. The cloud is supported by heavy industrial infrastructure, hidden from the user’s view.

6. Users see only the quick response, not the industrial chain behind it.

7. AI offers real benefits, but using it for low‑value content raises ethical questions in a world with finite resources.