The growing demand for AI data centers is raising concerns about their environmental impact and the potential for "digital extractivism," where richer countries exploit the resources of poorer nations to fuel their AI ambitions.
Data centers are energy-hungry, consuming huge amounts of electricity to power AI models. In fact, the training phase of large language models requires up to 30 times more power than traditional AI models. This has led to increased greenhouse gas emissions, with three of the biggest tech companies - Microsoft, Google, and Meta - reporting ballooning emissions since 2020.
Moreover, data centers require massive amounts of water for cooling, which can strain local resources, particularly in water-scarce regions. The extraction of critical minerals for hardware manufacturing also poses environmental concerns.
To avoid digital extractivism, strategies are needed to ensure that AI development serves local needs and ecosystems, rather than just richer countries. This includes developing policies and regulations that promote sustainable AI practices and encourage the use of renewable energy sources.
Some experts argue that AI can be a powerful tool in the fight against climate change, enabling more efficient use of resources and optimization of renewable energy systems. However, others caution that the benefits of AI must be weighed against its environmental costs, and that low-tech solutions may be more effective in reducing carbon emissions.
