Jensen Huang, the CEO of the chipmaker Nvidia, recently compared the development of artificial intelligence to surgery. Just as a surgeon must cut a patient to save them, he argues that society must endure significant environmental strain and increased energy consumption today to unlock AI's long-term potential to solve climate change. This perspective suggests that the immediate, heavy reliance on fossil fuels to power massive AI facilities is a necessary sacrifice for a more sustainable future.
WHAT'S HAPPENING
Artificial intelligence relies on massive, power-hungry buildings known as data centers. These are essentially warehouses filled with thousands of specialized computer chips—often manufactured by Nvidia—that work around the clock to process information. Because AI models require immense computing power, they need constant electricity. Currently, the demand for this power is outpacing the availability of clean, renewable energy, leading many companies to fall back on traditional fossil fuels like coal and natural gas. Some industry leaders argue that this spike in pollution is a temporary hurdle on the road to better technology that will eventually help us optimize energy use and fight climate change more effectively.
The hidden cost of computing
HOW IT WORKS
To understand why AI is so energy-intensive, think of these data centers as high-speed factories. The AI you use, such as a chatbot, is driven by a model—a complex digital engine trained on vast amounts of data to recognize patterns and generate responses. Running this engine requires two distinct phases. First, there is the training phase, where the model learns from massive datasets, requiring thousands of chips to run at full capacity for weeks or months. Second, there is the inference phase, where the finished model performs tasks for everyday users. Both phases generate immense heat and require consistent, round-the-clock electricity. When that electricity comes from burning fossil fuels, the environmental impact—in the form of greenhouse gas emissions and air pollution—happens immediately, even if the promised climate-saving benefits of the AI remain years away.
WHY IT MATTERS
The central tension here is a question of who bears the risk. While advocates argue for the future potential of the technology, the current reality involves immediate trade-offs. Relying on fossil fuels to power the AI boom can lead to increased local air pollution and further contribute to the warming that causes wildfires and sea-level rise. While some hope that AI will eventually manage power grids more efficiently or discover new materials for batteries, others point out that the energy choices we make today are not fixed. Governments and corporations are currently deciding whether to build new coal and gas plants to keep up with AI demand or to invest exclusively in wind, solar, and nuclear power. The path we choose determines whether the cost of AI progress is a temporary disruption or a long-term environmental burden.
Liked this one? The next lands at breakfast.
Every story in tomorrow's AI news, rebuilt in plain English — five minutes, sources linked, free forever.
By joining you agree to receive Article's daily newsletter — unsubscribe in one click. Privacy