US President Donald Trump signed an executive order on April 8 to boost coal production in a bid to cater to the rising power demand for setting up data centres in the US.
Through this move, which is counteractive to the global efforts to mitigate carbon emissions, Trump claimed that he was “taking historic action to help American workers, miners, families and consumers”, as reported by Bloomberg.
Trump also announced that the closed plants will be reopened if "they’re modern enough" or they would be demolished. He further shared plans to build new plants with the goal of “putting the miners back to work”.
Trump’s move is in line with his commitment to use America’s coal resources as a source of both electricity to operate data centres and heat to make steel. The US views boosting coal power as both a matter of national security and an asset in its bid to dominate the artificial intelligence industry.
Race to Dominate AI industry
A January 24 report by Reuters revealed that Trump announced a joint investment of up to $500 billion by ChatGPT creator OpenAI, SoftBank, Oracle and others to fund artificial intelligence infrastructure and to create over 100,000 jobs in the United States, in a bid to keep the US ahead of China in the race for dominance in artificial intelligence.
According to a report by Energy Information Administration, coal accounted for about 16% of US electricity generation in 2023, making it the fourth-highest energy source. This is a sharp drop from 51.4% in 2000, because the production of natural gas and renewable energy slashed dependence on coal for generating power. Data from Global Energy Monitor stated that about 770 individual coal-fired units have shuttered, as noted by Bloomberg.
AI: Propellant or Deterrent for a Cleaner Future
Artificial Intelligence (AI), as a tool, presents a dilemma for its users. On one hand it can assist in use cases that may lessen environmental impacts.
AI models due to their complexity , require considerable amount of computational power. To meet this demand, high energy is required especially during the training phase, which can be extensive for more sophisticated models, according to an EY report. If this energy is sourced from non-renewable power plants, it could cause greenhouse gas(GHG) emissions.
The massive infrastructure also requires cooling system that consume vast amounts of water to prevent overheating and sustain performance. This could put a strain to regions where availability of water is already scarce. Adding to the environmental concerns, the production and disposal of hardware underlying AI (including servers, GPUs and other specialised equipment) also has an environment impact.
On the other hand, it could accelerate environmental harm because of its growing energy consumption to keep its sophisticated models running, as per the EY report.
Although legislative frameworks are being established to integrate sustainability considerations into the development, operation and application of AI, ensuring companies can run their systems while complying with these standards, remains a challenge.
