AI craves data like water and energy.



AI’s craving for data is matched only by a runaway thirst for water and energy. The computing power for AI models requires immense natural resources, leading to significant energy consumption and water usage in data centres. Legislation is needed to address this potential environmental crisis.

Key Elements:

  • AI models require large amounts of natural resources for computing power
  • Data centres consume massive amounts of electricity and water for cooling systems
  • Energy and water usage by AI technologies is raising environmental concerns
  • Legislation is being proposed to address the environmental impact of AI

One of the most pernicious myths about digital technology is that it is weightless or immaterial, but in reality, AI technologies require significant natural resources. Data centres, which house servers for AI computations, consume large amounts of electricity and water for cooling systems. Legislation is needed to prevent an environmental crisis caused by the energy and water consumption of AI technologies.

The industrial feeding frenzy around AI is increasing the materiality of computing and raising concerns about energy consumption. AI models require specialized processors (GPUs) that are more complex and energy-intensive. Cooling systems for AI data centres also require substantial water usage, leading to environmental impacts.

OpenAI’s boss has warned of the energy crisis AI faces, with generative AI systems expected to consume vastly more power than previously thought. Efforts are being made to address these concerns, with proposed legislation in the US and Europe requiring AI systems to report their energy consumption and environmental impacts.