Blog image
Blog Post6 MINUTES

3 Sustainability Considerations You Need to Know with AI

PUBLISHEDSeptember 19th, 2024
SHARE

AI has had an increasing rapid uptake over the last 10 years, with one in six companies in the UK now using AI in one form or another, and for good reason: AI added an estimated £3.7 billion to the economy in the UK in 2022 alone. It’s safe to say we are now in the AI age, but it's not all generated rainbows. Nobody really knows how much energy AI is consuming, but researchers are starting to put estimates together, and it isn’t looking good. That means wider AI uptake leads to more energy needing to be generated, more parts to build more data centers, more chips to be manufactured with rare earth metals, more water used for cooling rather than drinking, and more hard-to-recycle waste. Not taking this seriously can cost you customers, your reputation, or lead to ESG failures, all of which can and will impact your bottom line.

Collective uptake of the technology has already happened, but individual companies can still, and are still, making strides in countering these problems, as well as facing some stumbles. Empowering your employees to combat these challenges through common sense training and effective use of DEX tools is a valuable new area of a growing industry to explore. Here are the key areas for concern, and some tips on how best to approach the rise of AI sustainably in your own workplace.

1. Energy Consumption

The computational power required to train and operate large AI models, such as those used in natural language processing, is immense. This energy consumption is a major contributor to carbon emissions, particularly when the data centers supporting these models rely on non-renewable energy sources.

Google has committed to achieving net-zero by 2030, but their emissions have increased 48% due to increased use of their AI service, Gemini, and corporate use of AI systems. Microsoft has made a similar claim, and had a similar setback. Nonrenewable power is still the majority of power produced globally, but carbon offset schemes and commitments to more renewable sources can and should be a major part of any businesses environmental strategy around AI.

These trends highlight the need for businesses to adopt and stick to Net-zero aspirations and invest in renewable sources for power generation, but AI can be used to increase efficiency and safeguard against its own impact with 2 simple steps:

  1. Only use AI where needed. AI can be the best solution to specific problems, but trying to utilize it everywhere is not only a bad business move, it’s also a strain on the environment. Try instead, to identify where AI is most needed in your business to increase efficiency or create better processes, and save more traditional processes where AI is not needed. AI can be your most powerful tool used in your workforce if it’s used strategically.
  2. Use AI to optimize software: Over 15% of the energy used in data centers can come from poorly optimized software. But using software optimization tools can reduce the energy load of your data centers, reducing energy consumption, heat generation, and water usage. You can read more about smart software license optimization here.

2. Water Usage

Water usage for AI is becoming a global concern. Water can be evaporated at a rate of up to 9L/kWh when used in data centers, depending on the climate and local water systems. 2024 has only seen an increase in droughts globally, when paired with water also being used for data centers this can lead to water rationing and shortages. Therefore, stewardship of water and responsible AI use go hand in hand.

Amazon Web Services has increased their efficiency of water usage by 50x; their use of recycled water, and invested in water replenishment efforts in areas where they operate data centers. Similarly, Microsoft is using innovations like submerged data centers and ambient water cooling from the ocean to make an impact (4). More individual and local solutions can include:

  1. Supporting companies that prioritize water efficiency and transparency in their operations. Look for platforms and services that disclose their sustainability efforts, including water stewardship initiatives. Strong public partnerships between businesses and stewardship efforts can help raise your environmental profile in the public view, and can be a major factor for environmentally conscious talent you are looking to acquire.
  2. Ensuring the rest of your workplace is water efficient, to minimize the draw on the local system. Set dishwashers, toilets, taps and other standard office appliances to run less frequently and more efficiently with simple changes. Reminders through DEX tools like Engage can be a great help in this area.

E-Waste

Nvidia became the most valuable company in the world in Summer 2024, valued at over $3 Trillion at its peak. The reason is simple: the massive uptake of generative AI technologies and their hardware requirements. The chip producer manufactures millions of chips per year, and is far from the only hardware supplier involved in the AI logistics supply chain. These chips can last anywhere from three to ten years and will need replacing. Each chip is made of many rare earth metals including amounts of silicon, copper, boron, gold, and other elements. Recycling these components is both an environmental and economic concern.

So what can businesses do? Well, Apple and Dell offer good examples. Apple introduced ‘Daisy’—a robot designed to break down and recycle iPhones to recover valuable metals and other materials, whereas Dell offers a recycling service where old equipment can be sent to them for disassembly and recycling. Other practical steps that can be taken to reduce e-waste include:

  1. Ensuring your company operates or takes part in a recycling initiative, where obsolete equipment is properly recycled or sent to facilities that can do so. Check with your local recycling services on where you and your employees can bring E-waste.
  2. Extending the lifespan of your equipment through regular maintenance, only upgrading when needed, and operating equipment in its optimum configurations. In a recent study by Nexthink, researchers found that over 90% of modern laptops are Windows 11 compatible—thus eliminating the need for a new hardware replacement.

Above all, employee engagement in these strategies is paramount to their effectiveness, and investing in a strong DEX tool is vital. With dispersed workforces and less frequent office access, updating your teams on tech issues, new policies, and regulation changes can mean the difference between significant energy waste or unprecedented savings.

Request a Nexthink Demo