Deutsch English
  • Deutsch English
  • Tech of Tomorrow
  • Morals & Machines
  • Burn to Learn
  • Think Tanks
  • About us
  • 22.12.2021
  • Mitzi László

4 Ways Big Tech Could Reduce Its Carbon Footprint

Being online is energy intensive, but there are simple steps Big Tech could take to reduce its environmental impact.

Large buildings full of servers humming away being fanned to decrease the temperature. Submarine cables that run under oceans. Pause. Think about that. Someone put a cable under the ocean. It then connects to yet another hot building full of blinking lights. There is a lot of stuff that requires a lot of energy to store data, send data, train algorithms, etc.

The average person does not see the equipment that makes the Internet work on a day-to- day basis. Information appears on a screen without being physically plugged in making it easy to forget that there are vast amounts of physical machinery powering it. In combination with a greenwashing marketing machine from Big Tech, it is easy to get the impression that digitally stored data uses little or even no energy.

The energy costs of meeting online with live high-quality video are even hailed as saving energy. Meanwhile, flight shame is rampant as we watch aeroplanes spit out trails across the sky. What’s the real difference in the amount of energy consumed between an in-person meeting, with all the embedded transport costs, and holding an online meeting?

It is estimated that Big Tech is responsible for at least 110 million metric tons of CO2 emissions per year

There are different ways of defining the amount of energy consumed and emissions emitted by Big Tech. It all depends on the precise details. Estimates have been made. I’ve attempted to piece together data to get a very rough picture of the quantity of CO2 emissions by Big Tech.

Alphabet (formally Google), Apple, Meta (formally Facebook), Amazon, and Microsoft (collectively known as GAFAM) all produce annual sustainability reports with carefully padded data about the amount of CO2 emissions each claims responsibility for. Even these five corporates take very different approaches to how they assess their environmental impact and the measures they take to offset it. Last year, according to their own reports, the largest five tech companies were collectively responsible for at least 110 million metric tons of CO2. To put that in perspective, the whole planet emitted 34.8 billion metric tons of CO2 over the same year.

There are approximately 1.4 billion Chinese people where BAXT (Baidu, Alibaba, Tencent, and Xiaomi) dominates the market and roughly 1 billion people in the United States and Europe where GAFAM dominates the market. Based on population data, we could make the (very) broad assumption that there are a roughly similar number of users of GAFAM and BAXT. Therefore, the CO2 emissions could be roughly comparable.

Again, these are rough estimates; the point is not to haggle over the forensic details. What we do know is that Big Tech’s CO2 emissions are significant.

4 ways Big Tech could reduce CO2 emissions

The sustainability reports of Big Tech all focus on using cleaner energy sources like solar and wind power, which is a step that can be taken retroactively. 

It is possible to design computational infrastructure in such a way that it uses less energy and emits less CO2 in the first place. Here are some concrete suggestions for how Big Tech could contribute to reducing CO2 emissions. 

1. Build repairable digital equipment that lasts

Digital equipment includes not only the running of the computational infrastructure but also the production of the physical objects used in computing. This includes computers, mobile phones, and tablets as well as databases and submarine cables. Digital infrastructure relies on rare metals and water as well as energy to be built.

The current lifespan of many devices is fairly short, with many having programmed obsolesce, meaning they are programmed to break, or no effort is made to maintain older devices, forcing users to throw away physical equipment and purchase newer versions. 

Devices should be repairable and be made to last rather than have inbuilt redundancy. Maximizing the lifespan of resources would reduce the need for more devices and make the most of what we’ve got. Rather than working our way to the iPhone 100, there needs to be an iPhone For Life that you can treasure, take care of, and repair, and when something breaks, it should be able to be recycled. Software updates need to include the design principle of keeping older devices active rather than pushing for newer features we don’t really need that only work on new devices.

FairPhone is leading the way in proving that this is possible. If the 200 million iPhones sold per year adopted the FairPhone model, we would be a big step closer to completing the sustainability transition. The same principle goes for Android phones, as well as all other digital devices, such as tablets, computers, databases, and cables. 

Today, Agbogbloshie, a suburb near Accra in Ghana, is famous for its collection of global e-waste. It is possible to avoid such disasters if Apple and Google, in particular, committed to building digital devices that were built to last and that could be repaired.

2. Only collect necessary data

It is important to be more conscientious about the data we collect and store. Big Tech relies on big data, and an ever-increasing amount of data, rather than smart and intentional data collection. Big Tech collects large amounts of data in the hope that it will one day be useful. Yet data hoarding is energy intensive and, in many cases, unnecessary. It’s time for Big Tech to Marie Kondo their data collection practices. 

With the Internet of Things, sensors are appearing left, right, and center around the home and public spaces. It’s becoming harder and harder to buy household appliances without sensors. Each of these sensors communicates with a storage system and generates data. All of this requires energy, not only to run the device but to send the data to a database that is then contributing to models and analytics.

Article 14 of the General Data Protection Regulation talks about ‘data minimization.’ If Big Tech committed to only collecting data that produced real value, it would be contributing to the sustainability transition. Big Tech taking more care to think about what data is beneficial to collect and why would go a long way to improving its sustainability credentials.

3. Store data locally

Why send everything to Silicon Valley? If you look at a map of submarine cables, it is remarkable how many paths lead to the USA and Europe. Centralizing data is not technically necessary, but it is the result of how global power is structured. The energy costs of centralizing power are substantially more than carrying out computing more locally. 

Data can be stored physically closer to where it is being used. When we store data locally, it does not need to travel so far and therefore uses less energy. If Big Tech, particularly Amazon and their Amazon Web Services offering, committed to keeping local data local, it would contribute to the sustainability transition. 

4. Train Artificial Intelligence intelligently

Timnit Gebru was fired from Google over a paper that in part discussed how vast energy consumption by Artificial Intelligence (AI) would result in inevitable bias, as only those who had access to that much energy would be part of training the AI. The training procedure of a particular AI model emitted 284 tons of CO2, which is equivalent to the energy involved in a trans-American flight. And that’s just the energy involved in training. To put things in perspective, an average person emits roughly 5 tons of CO2 annually. 

Training algorithms needs to be done sparingly and intentionally. As with data minimization, Big Tech could contribute to the sustainability transition by training fewer AI models and only doing so when there is a clear benefit to society. Ideally, society would have a voice in AI and its structure considering the amount of energy it uses and the impact it has on our collective understanding of the world.

Who’s responsible for Big Tech?

There are relatively few people who have significant sway over the nine Big Tech companies. 

Larry Page and Sergey Brin collectively own 51.2% of Alphabet, the parent company of Google. Mark Zuckerberg owns 55% of voting shares in Facebook, recently rebranded to Meta. Amazon is largely controlled by Jeff Bezos, who owns 11.1% of shares but only needs 33% of the shareholders to side with him in any proxy vote. BlackRock Inc., with CEO Larry Fink, owns 5.4% of Amazon, but they also own 6.6% of Apple and 6.8% of Microsoft. Vanguard Group Inc., with CEO Mortimer J. Buckley, also owns 6.6%, 7.83%, and 8.4% of Amazon, Apple, and Microsoft, respectively.

Half of Apple is owned by other corporations, with 20% owned by BlackRock Inc., Vanguard Group Inc., and Berkshire Hathaway Inc. Similarly, almost 19.4% of Microsoft is owned by three institutions, two of which are the same that own Apple, i.e., Vanguard Group Inc., BlackRock Inc., and State Street Corp. Basically, very few people are responsible for deciding how Big Tech operates.

Power mapping the individuals behind corporates is a helpful exercise to understand how many people need to agree to making a change. While international politics requires consensus between large numbers of people, corporations, especially those in Big Tech, have relatively simple power structures, with only a few people having a very influential voice.

An active choice

Each of these suggestions is in direct opposition to the current business models of Big Tech. While it’s easy to switch to wind and solar energy, the real impact can be made by thinking through how to make the most of our resources and how to design sectors to be more efficient. This will require a shift in priorities by the corporate elite. It is possible to think of new business models that would still be profitable. Perhaps not as profitable, but Big Tech doesn’t need as much profit as it has for all shareholders and employees to be paid enough to live a good life. 

Additionally, many of the energy efficiency solutions for Big Tech will also have a positive impact on data protection and cybersecurity. Long lasting devices, more intentional data collection and use, as well as locally governed data storage, would also have a positive impact on human rights. So, tackling Big Tech’s energy use could present a wonderful opportunity for ‘hitting two birds with one stone,’ and creating real change across several sectors of society.

Big Tech has enormous power over the state of affairs of digital rights, including reducing global CO2 emissions. Nine companies, controlled by only a handful of people, could make a dramatic contribution to the global energy transition.

Data on CO2 emissions is usually given per capita rather than per company — a signal of how emissions narratives place blame on the individual rather than on corporates. The meager reduction in CO2 emissions during the pandemic lockdowns demonstrated that even if we all reduce our individual carbon emissions, it is not nearly enough to reach the Paris Agreement targets, limiting global warming to 1.5 degrees. The only way to maintain a habitable planet is if the people behind the corporates prioritize reducing CO2 emissions over power concentration.

With the knowledge we have today, it is impossible to say that those controlling Big Tech are just going with the flow. No, they are actively choosing not to prioritize the sustainability transition over power concentration, and they hold the responsibility for the consequences of that decision. If the few people controlling Big Tech chose to switch gears and prioritize the habitability of the planet and the wellbeing of society, we could see real change.

Cover Image: Getty

Mitzi László

Mitzi László is an advocate for equity enhancing computational infrastructure based in Amsterdam. She focuses on the impact of governance between multiple parties and political economies in determining the degree of privacy, transparency, fairness, data protection, and distributed computing. Previously, she worked for Sir Tim Berners-Lee’s startup, Inrupt/Solid, aiming to improve personal data control. Mitzi László also works as an independent ethical evaluator for the European Commission as well as for Next Generation Internet Pointer.

You might also like

  • Think Tank
  • 13.05.2022

The end of thinking as we know it

If we want to keep up to speed in an increasingly complex world, we know that we need to enrich and expand our human intelligence. We can do that with the help of artificial intelligence.

  • Think Tank
  • 28.12.2021

The Club of Living Romantics

What we are experiencing in Germany today is a reverse disruption.

  • Automation
  • 28.07.2021

Time to replace humans

People are terribly inconvenient. They pride themselves on working a lot but are obscenely inefficient. A call for automation.

© 2022 ada
Imprint
privacy policy