The latest episode of the Greenpeace International podcast System Shift dives into one of the biggest climate blind spots of AI: the energy hunger behind AI chipmaking.
Why is AI bad for the environment?
You may have heard about AI’s astronomical energy use of data centres, its dependence on fossil fuels and resulting greenhouse gas emissions.
But that’s only part of the story of why AI is bad news for the environment.
The energy hunger behind AI chipmaking
The US company NVIDIA, owner of ChatGPT, made history last month when it became the first company to reach $5 trillion in market value. NVIDIA promotes its products as ‘sustainable computing’ and a climate solution, yet its supply chain tells a different story.
Manufacturing AI hardware is energy intensive. While the company is based in California, many of its major suppliers are actually operating factories in East Asia, in countries such as Korea and Taiwan.
NVIDIA is essentially outsourcing its pollution to a region already heavily dependent on fossil fuels and deeply exposed to climate risks.
Join the conversation with Greenpeace East Asia
Join the conversation with Katrin Wu, Sunryul Kim, and Lena Chang from Greenpeace East Asia and find out about their latest research into how AI is bad for the environment and what true clean, responsible AI should look like.
