As artificial intelligence (AI) and cloud computing power our digital world, their environmental costs are overshadowing the once-optimistic promises of efficiency and innovation of the industry. AI and cloud technologies are gradually contributing to a growing energy crisis even though it is being hailed as transformative forces healthcare, finance, communication and more such industries.
Microsoft’s rumored $100 billion “Stargate” project is a proposed AI-focused data center in Wisconsin. It is learned to consume a five gigawatts of electricity and this is enough to power about 4 million homes. It is just one example on how AI-driven infrastructure demands immense energy resources.
The tech sector has played an important role in reducing energy consumption through more efficient data processing and hardware innovations for years. The new trend is something different and opposite. A recent report from Grid Strategies highlights that the energy demand in the U.S. is set to rise by 5% over the period of next five years and it is mainly driven by the explosive growth of AI-powered data centers.
The problem goes beyond the mere energy usage of the data centers. The fundamental design of AI systems such as machine learning models like OpenAI’s ChatGPT is energy-intensive. The complex computations required to train and run the systems generate massive heat and therefore leads to even higher energy costs for cooling. Around 40% of a typical data center’s electricity bill is allocated to simply keeping its servers from overheating.
It is the right time to acknowledge the environmental damage that the companies are driving. AI was meant to help solve global challenges, but its power demands have been continuously unchecked and we risk worsening one of humanity’s greatest threats—climate change.
Michelle Solomon, analyst at Energy Innovation, put it as a “litmus test” for a society that is facing climate change. It is now being questioned that whether we should allow the insatiable energy consumption of AI to justify extending the life of coal plants and building new gas facilities.
