High carbon emissions are the hidden cost of AI expansion

High carbon emissions are the hidden cost of AI expansion
Artificial intelligence is rapidly transforming our world, but it comes at a hidden cost: high carbon emissions

Artificial intelligence can process data at lightning speed and the latest version of ChatGPT can now converse with users and identify images. While growth in AI and Cloud computing services has been transformative in a variety of sectors, it is straining data centers and electricity grids because of the huge amounts of energy consumed.

The UAE and Saudi Arabia are among Gulf states that have adopted artificial intelligence and digitalization in government, business and industry. Their advantage lies in modern infrastructure, unburdened by legacy systems that could hinder their net-zero ambitions.

Setting aside debates on the benefits and threats of game-changing technological advances, the carbon footprint of AI and Cloud computing services needs to be managed. The challenge lies in addressing this surge in electricity demand in a way that is sustainable and does not conflict with climate change goals. Developers like Microsoft, Google and OpenAI are all working on solutions, including using more renewable energy and more sustainable cooling systems.

While the global electricity system is shifting with the rapid growth in renewable energy, clean energy sources – solar, wind and hydropower – make up just 30 percent of global electricity generation. Oil, gas and coal still dominate, accounting for more than 80 percent of primary energy consumption. A number of studies indicate that the surge in demand from AI and cloud computing has led to higher carbon emissions, undermining decarbonization efforts. A Columbia Climate School paper titled  “AI’s Growing Carbon Footprint” estimates that the greenhouse gas emissions from the energy consumed by data centers exceed those of the aviation industry.

Data centers have witnessed a higher volume of online activity since the 2020 coronavirus pandemic, when remote working habits and curtailed travel meant more businesses went virtual. The growth in the use of smart phone applications, bitcoin mining and crypto-currency transactions has also contributed to higher demand for energy and data.

The processing of data generates heat with each “bit” change so data centers need cooling at all times, adding further to the electricity load.

At the Petro Diplomacy conference organized by the Arab Gulf States Institute in Washington in June 2023, attendees heard that bitcoin mining uses the electricity equivalent to the entire diesel fuel consumption of the U.S. rail fleet. The conference also heard that the energy used to train the ChatGPT3 data application – already superseded by a more advanced version – was equal to the energy use of 138 U.S. homes in one year.

The Columbia report says a single request in ChatGPT can consume 100 times more energy than one Google search. A single AI training module can consume more electricity that the annual usage of 100 US homes, according to a Bloomberg report titled “Artificial Intelligence Is Booming—So Is Its Carbon Footprint.”

Sara Vakhshouri, CEO of SVB Energy International, noted that “the carbon dioxide emissions generated from this training were equivalent to three round-trip flights from New York City to San Francisco.”

Regarding cryptocurrency, the Columbia Climate School refers to the “dirty little secret” that is associated with the virtual currency. It estimates that bitcoin consumes an estimated 150 terawatt-hours of electricity annually, more than the entire consumption of Argentina, emitting 65 megatons of carbon dioxide into the atmosphere annually.

These rapid advancements in technology make it difficult to quantify the impact of digitalization, AI and cloud management applications on energy demand. Most forward-looking forecasts tend to base their modelling on population and economic growth trends. Oil demand forecasts are based on the number of expected electric vehicle sales, heat pumps and other applications, most of which rely on electricity. And those EVs need to be charged, which increases demand for electricity.

The IEA said in presenting its net zero report that achieving net zero carbon emissions by 2050 “will require large new and smarter infrastructure,” including electricity transmission and distribution grids that “need to expand by around 2 million kilometers per year to 2030.”

The UAE, which will host the COP28 climate summit at the end of November, has one of the Gulf region’s most ambitious net zero targets. It is investing in digitalization, smart grids and AI to manage its oil and gas operations.

Abu Dhabi investment vehicle ADQ, which was formed in 2018, is working on critical AI projects in oil and gas, such as drilling performance, reservoir modelling, corrosion detection and product quality.

Read more: The GCC is serious about carbon capture

Saudi Aramco is also betting on artificial intelligence to enhance its operations. Facilities in remote desert sites or at sea present a serious data transmission challenge, so the oil giant is looking to use a new generation of satellites to gather data. The Kingdom is also scaling up renewable energy capacity and plans to spend more than SAR1 trillion ($270 billion) by 2030 to boost Saudi clean energy generation capacity, as well as to expand and modernize its transmission and distribution networks, Energy Minister Prince Abdulaziz Bin Salman Al Saud said at the IKTVA 2023 Forum in Dhahran in February.

This will require investment in new data cloud centers and more are planned both in Saudi Arabia and in the UAE. Microsoft said in February that it would be investing in a new cloud data center in Saudi Arabia while Oracle plans to invest $1.5 billion to meet cloud computing demand in the Kingdom at a new base in Riyadh. It already has a cloud region in the coastal city of Jeddah and plans an additional facility in the futuristic city of Neom, the company said in February.

Both Saudi Arabia and the UAE are betting on growth in cryptocurrency although that too has implications for carbon emissions due to the amount of energy needed to process virtual currency transactions.

While a new, less carbon intensive energy landscape is taking shape, it will take time to scale up. More investment is needed to stabilize and adapt existing grid infrastructure to accommodate intermittent energy sources. A modernized, digitalized grid would support wider deployment of renewable energy and enable rationalization of energy consumption through inter-connectivity and data sharing but this will require strong policy and regulatory measures, some of which are still not in place in many parts of the world. These will need to be accelerated to keep pace with the rapid advances in cyberspace.

For more news on energy, click here



Disclaimer: Opinions conveyed in this article are solely those of the author. The information presented in this article is intended for informational purposes only. It does not constitute advice on tax and legal matters; neither are they financial or investment recommendations. Refer to our full disclaimer policy here.