By: Ben Selier - Vice President, Secure Power, Anglophone Africa at Schneider Electric

There’s isn’t a day, let alone an hour that goes by that AI isn’t featured in a new headline or newsletter update. AI has become a new team member or research assistant of sorts; however, this overachieving mate of ours also comes with environmental consequence.

Like DC superhero, The Flash, AI is extraordinarily fast and intelligent but also consumes and inordinate amount of energy.  The result - rising energy demand and a resultant carbon footprint which go head-to-head with globe’s sustainability goals.

Power like no other

The computational power behind AI is massive. OpenAI researchers Dario Amodei and Danny Hernandez observed that from 2012 to 2018, the compute required for deep learning research doubled every 3.4 months—a 300,000-fold increase in just six years which far exceeds Moore’s Law, which states that processing power doubles every two years.

However, and this is good news, as the world focuses on climate change, AI researchers are also beginning to recognise its carbon cost. A study by Roy Schwartz et al. at the Allen Institute for AI questions whether efficiency, along with accuracy, should become a priority. AI models require vast amounts of computational power for training data processing and experimentation, which drives up carbon emissions.

Similarly, the University of Massachusetts (Strubell et al., 2019) highlighted the environmental impact of AI, analysing the computational demands of neural architecture searches for machine translation.  

Therefore, five years ago already, it was projected the carbon cost of training such models is at 626,155 lbs of CO₂, equivalent to 125 round-trip flights from New York to Beijing. As AI's energy demands continue to grow, it's vital to consider sustainability alongside utility.

The good news is that like any superhero, the AI also provides significant benefits which includes sustainability. A joint study by Microsoft and PwC in 2019 projected that responsible AI use could reduce global greenhouse gas (GHG) emissions by 2.4 gigatonnes (4%) by 2030.

As an example, Google has employed machine learning (ML) from its DeepMind division to optimise energy usage in its data centres, achieving a 35% improvement in energy efficiency. AI is also enabling precision agriculture, where predictive algorithms minimise water and fertiliser use, improve crop yields, and reduce waste. In renewable energy, AI plays a fundamental role in forecasting solar and wind energy output, ensuring these natural resources are used efficiently.

Smarter data centres

As demand for AI applications grows, so does the reliance on data centres, which are increasingly critical to powering ML and deep learning models. However, data centres already account for a significant portion of global ICT energy use. According to the Shift Project, the ICT sector accounts for around 4% of global carbon emissions, with its contribution to GHG surpassing that of the aviation industry by 60%. 

Furthermore, as more businesses adopt AI to drive innovation, the demand for cloud-optimised data centre facilities will rise. By 2025, data centres will account for 33% of global ICT electricity consumption. 

To minimise their carbon footprint, companies must ensure their data centres are equipped to handle high-density compute demands efficiently. Unfortunately, up to 61% of systems run by corporate data centres are running at low efficiency, says  ScienceDirect.

Sustainability begins with smarter data centre choices. Cooling alone constitutes about 40% of a traditional data centre's energy use, and in warmer climates, these costs are even higher. Indeed, data centres are expensive, especially in warmer climates, and more than 80% of hardware does not need to be near the end user in terms of latency

Companies like Google have demonstrated the advantages of locating data centres in cooler climates, such as Nordic countries, where natural cooling can dramatically reduce energy consumption. Iceland, for example, offers an ideal combination of low temperatures and 100% renewable geothermal and hydropower energy.

The future of AI can be sustainable

Sustainability in AI is not just about reducing its footprint; it’s about rethinking the infrastructure and processes that support its growth. Innovations such as liquid cooling are making data centres more efficient by directly managing the heat generated by high-performance computing. This technology is becoming an essential part of sustainable design for the future of AI.

Organisations must also employ a proactive and strategic approach to deploying AI responsibly. The World Economic Forum World Economic Forum (WEF) suggests a four-step process to balance AI’s benefits with its environmental impact:

1. Select the right use case: Not all AI optimisations lead to significant carbon reductions. Organisations should prioritise processes that can be meaningfully optimised by AI, especially for sustainability use cases.

2. Choose the right algorithm: The energy consumption of an AI system depends largely on the algorithm used. By selecting the most efficient algorithm, organisations can significantly reduce training time and energy usage.

3. Predict and track carbon outcomes: Good intentions alone aren’t enough. AI implementers must include carbon footprint estimates in cost-benefit analyses and use sustainability as a key performance indicator for AI projects.

4. Offset the footprint with renewable energy: Organisations must utilise green energy sources to power AI models. Google has committed to powering its data centres with renewable energy, achieving net-zero carbon emissions since 2017.

AI holds extraordinary potential to improve lives, drive economic growth, and tackle global challenges. But its future depends on ensuring that the infrastructure and energy systems supporting it are both efficient and sustainable.