In contrast, models from companies like OpenAI are known for consuming vast amounts of energy for data processing, which in turn increases the demand for fossil fuels, especially natural gas, to power global data centers.
DeepSeek’s energy efficiency could mean that the expected exponential growth in energy demand at data centers might not materialize as anticipated.
The oil and gas sector, which has seen static demand for some time, had hoped that AI-driven sectors would increase the need for fossil fuels to power advanced computing systems.
However, if AI operations prove more energy-efficient than expected, these companies might need to seek alternative markets for their products.
Should the energy demands from AI not escalate as projected, this could lead to a reevaluation of environmental policies and regulations concerning energy production. There might be less impetus for new fossil fuel projects at a time when the industry can least afford such a shift.
As DeepSeek’s claims of energy efficiency make headlines, there’s pressure on other AI developers to adopt similar efficient practices, which could alter the global energy demand landscape. This shift might reduce the strain on oil and gas resources, potentially impacting fossil fuel prices.
Regarding projected energy demands, significant figures have been cited: Wells Fargo, for instance, did forecast in 2024 that AI power demand could rise from 8 TWh to 652 TWh within six years, an 80-fold increase.
