Bill Gates, in his book ‘How to Avoid a Climate Disaster’, highlights the grave reality our world faces: the production of 52 billion tons of greenhouse gas emissions annually.
Bill Gates, in his book ‘How to Avoid a Climate Disaster’, highlights the grave reality our world faces: the production of 52 billion tons of greenhouse gas emissions annually.
While this situation is alarming and expected to worsen in the coming years, Gates contends that with the right tools – particularly the adoption of advanced digital technologies – we can progress towards achieving zero emissions.
Although technologies are effective in reducing carbon emissions, they come with their own environmental costs -- the digital carbon footprint. This refers to the total greenhouse gas emissions generated by the use of digital technologies comprising data storage, data processing, as well as the energy consumed by networks and devices. A growing reliance on technology is becoming a double-edged sword in our fight against climate change. This is a conundrum that cannot be ignored.
Artificial Intelligence (AI), for example, hailed as a game-changer presents a green dilemma. Although it is indispensable in tackling climate change, training and operating complex AI models and datasets require immense amounts of energy.
To illustrate this further, consider Chat GPT, a well-known AI model, which produces approximately 8.4 tons of carbon dioxide annually – double the amount emitted by an individual (4 tons). This is because training such large language models can consume thousands of megawatt-hours of electricity resulting in emissions of hundreds of tons of carbon.
Researchers at the University of California found that Microsoft consumed 700,000 liters of freshwater during the training of Chat GPT-3 at its data centers. This is because large amounts of energy are used, which is then transformed into heat. To counter this and keep temperatures under control, a huge quantity of freshwater is required, which is evaporated during the process to cool down the machinery.
In addition to this, the use of social media leaves a significant carbon footprint. TikTok, for instance, has one of the highest carbon impacts among all digital platforms. While social media, undoubtedly, has transformed the ways we communicate and interact with one another, our daily activities – such as search queries, cloud computing, attending virtual meetings, sending emails, online shopping, streaming our favorite shows, or simply scrolling through our socials – contribute to carbon emissions. Every click, like and message carries a digital carbon footprint.
As key drivers of online trends, influencers hold the power to shape consumer behaviour by raising awareness or launching digital minimalism campaigns encouraging mindful online activities
Facebook remains the top social platform, reaching 3.03 billion users worldwide, followed by YouTube, WhatsApp, and Instagram with each having over 2 billion users, while TikTok exceeds 1 billion. On average, people spend around 2 hours 23 minutes per day on social media.
Younger generations, influencers, and content creators leverage these platforms to engage with vast audiences. Their prime motivations may comprise social interaction, entertainment as well as information seeking, however excessively engaging in activities like digital consumerism, filming travel vlogs, live streaming or constantly uploading high-frequency videos impact the environment as these amplify energy demands.
While estimating the exact scale of digital carbon emissions remains challenging, studies do suggest that these could range between 2.3 per cent and 3.7 per cent of the total global carbon emissions, a figure comparable to that of the entire aviation industry. By 2040, it is anticipated that the Information and Communication Technological (ICT) sector will contribute 14 per cent of global emissions, with communication networks and data centres being the prime drivers.
In the wake of this, it is becoming essential to explore pragmatic solutions to minimize the digital carbon footprint while continuing to benefit from technological advancements. First of all, developers and researchers need to prioritize creating AI models that are less energy-intensive. This can be achieved by optimizing hardware and ensuring that AI systems require fewer computational resources without compromising performance.
Government and technology hubs should focus on investing in energy-efficient data centres, such as developing green data centers that are powered by renewable energy sources like hydropower, solar, and wind.
Second, it is significant to raise awareness about the environmental impact of the excessive use of social media. As key drivers of online trends, influencers hold the power to shape consumer behaviour by raising awareness or launching digital minimalism campaigns encouraging mindful online activities.
In addition to this, regulatory authorities also need to play a pivotal role in establishing and implementing guidelines that promote sustainable practices and accountability for all the relevant stakeholders.
The writer is a research associate at the Graduate Institute of Development Studies, Lahore School of Economics.
