On a train or bus, or just standing in a queue, essentially the most common sight lately is the dim glow of a screen and the flickering thumbs of individuals lost in infinite scrolling on their smartphones.
Around 62% of individuals worldwide are energetic social media users. In some countries this number is the case over 90%. This results in high usage: the typical adult within the UK spends money 3 hours and 41 minutes online each day, which corresponds to about 56 days a 12 months, or almost two whole months.
Every time we read an article, see an commercial, watch a photograph or video, this content should be transferred from the social media platform's servers to our device. The larger the file, the more data must be transferred. And high-resolution images or long videos generate lots of data.
This data is distributed across many “server farms” (often housed in a big warehouse with hundreds of computers) all over the world. When you load a video from YouTube, you aren’t connecting to a single “YouTube data headquarters” somewhere in California, but as an alternative are collecting data from many various servers, often in several countries or continents.
Transmitting data over the Internet requires energy and sends signals through various electronic devices, including routers, servers, and our own cell phones or laptops. Any of those devices consumes energy to operate while servers must be kept cool. And this energy is usually derived from fossil fuels.
According to a study conducted by Greenspector in 2021 and beyond amongst Internet users in France, Tiktok is the least environmentally friendly social media platform updated in 2023.
Simply scrolling through the app exchanges lots of data because Tiktok is continually playing videos, including many pre-installed within the background that you might never see.
At the tip of the spectrum is LinkedIn. As a text-based platform with fewer photos and videos, scrolling through LinkedIn uses much less data.
Generative AI is energy hungry
Of course, social media isn't the one wrongdoer. Generative AI, with its ability to create text, images, music and even videos, is totally transforming many creative processes. But while it's tempting and sometimes mandatory, it also comes at a value to the environment.
Not surprisingly, the more powerful AI is, the more energy it consumes. Unlike streaming video or loading a big webpage, generative AI uses most of its energy at the tip, while your request is being processed. If you ask ChatGPT to write down you a novel, the writing process requires lots of calculations, even when the resulting text itself doesn't use much data.
Of course, all of this raises critical questions on the sustainability of generative AI and our own carbon footprint. The AI firms themselves are reluctant to inform us how much energy they use, but they’ll't appear to stop their very own chatbots from taking successful. I asked ChatGPT-4, “How much energy was used to process this request?” and it said “0.002 to 0.02 kWh,” which can be “much like lighting a 60-watt light bulb for about two minutes “leave it switched on”.
This roughly corresponds to the figures offered by independent evaluation and uses ten times more energy than is required for a Google search. With hundreds of thousands of requests per day to ChatGPT alone, this all adds as much as an enormous amount of additional energy consumption. As generative AI continues to evolve, energy demands will only increase.
What you’ll be able to do
While the environmental impact of those technologies raises legitimate concerns, it is usually essential to acknowledge their advantages. To give an example, AI-powered tools like text-to-speech, speech recognition, and automatic captioning have already made society more inclusive, especially for disabled or neurodiverse people. I’m not suggesting that we do away with social media or reject generative AI altogether.
But there are things we are able to do to cut back the carbon footprint of our web use, combining individual actions and systemic changes. Here are some strategies we are able to all use:
First, limit screen time. That's essentially the most obvious thing. Reducing the time spent on social media can directly reduce energy consumption.
Second, use power saving settings in your devices, similar to: Such as reducing screen brightness, using a dark background, and enabling power saving modes.
Third, consider choosing less energy-intensive social media using environmental rating information to guide your decision. That means more text and fewer video and generative AI.
Fourth, use Wi-Fi over 4G or 5G cellular data at any time when possible: Wi-Fi generally uses less energy.
So next time we discover ourselves scrolling through infinite sequences of images and videos with our faces illuminated by the blue glow of our screens, let's just pause for a moment and begin implementing these easy strategies in order that at the identical time can enjoy the advantages of connection while minimizing the impact on our planet's resources. Ultimately the alternative is ours.