Offsetting vs. Innovation: The Debate on How to Make Tech More Sustainable
Can it be done?
There are many different ways to make tech greener, hopefully. The challenge will be establishing a clear understanding of what actually makes something greener.
Without reducing the output, the rates at which offsetting must continue will be set by the output. The tail wagging the dog.
One question in particular that will gain momentum by 2025 is “Is offsetting enough?”
Why does it need to happen?
Digital advancements impact each other and technology is constantly trying to catch up with itself. Streaming, for example, wouldn't be as possible or popular as it is today without the mass adoptiontion of smartphones.
Now, nearly everyone has a 24/7 convenience store and festival of entertainment in their pocket. People use the internet every single day for both work and pleasure, and as we are all privy to, this is only increasing.
All this usage has resulted in the internet becoming “the largest coal powered machine” according to the The Green Web Foundation
What do we already know?
Technology is built on the foundation of always or soon to be on. It needs a constant supply of energy; where that energy comes from is a good place to start in the “how can we make tech greener” debate?
Over the years, demand has continued to increase on Data Centres but thankfully, we have also seen a substantial increase in energy efficiency and cooling methods. This has “helped to limit growth in energy demand from data centres globally”, according to intergovernmental organisation IEA (International Energy Agency).
Where are we heading?
Though, as the IEA also points out, data centres and data transmission networks are still responsible for nearly 1% of energy-related GHG emissions.
In a VMWare OCTO blog (The OCTO blog delivers unique perspectives on VMware's technology direction) Engineering Director Gavin Lu wrote; “the world’s sustainable economy calls for data centres to reduce emissions while boosting their capacities.”
Currently, we’re seeing an increased focus on climate goals generally but there is a pointed focus on the relationship between tech and the climate.
What does this look like in real terms? Bettering the current tools and technologies used and increasing their efficiency. Using other technology to offset energy usage. And inventing new approaches for future solutions that will be built with sustainability and longevity in mind.
About a year ago, McKinsey published an article titled ‘Delivering the climate technologies needed for net zero’ and within the article it outlined the 10 families of climate technologies that can play important parts in mitigating carbon emissions.
...And what about AI?
It depends on who you ask, but AI is either the answer to everything or it’s another bit of tech we’re going to need to offset in the future.
To go some way to avoid the latter, efficiency and sustainability must be established at every stage. This includes everything from the sourcing of sustainable materials to build the hardware, to choosing an efficient algorithm and using model compression to decrease size but retain complexity, to using renewable energy resources to power the machine in its learning stage and its more intelligent self.
With a tech-for-good head on, this all looks like great news; there is a way to make ML/AI as sustainable!
But with them being so new and the data surrounding the new AI tools available to the public being shrouded in mystery by their creators; we don’t yet know the real impact these tools are having.
Take ChatGPT for example; a recent estimate put its energy consumption in January 2023 alone as the equivalent of 175,000 Danes. For the complex method of working this out, take a look at this article published by Towards Data Science.
The future is in our hands
People’s desire for ever more efficient technology intersecting with people's desire to ensure less damage to the environment puts us at an interesting point in technological advancement. Tech-for-good as a movement is gaining momentum, but so too is the incredibly resource heavy AI industry.
Ultimately, achieving a greener tech industry will require ongoing efforts to prioritise sustainability and develop new approaches for future solutions. While progress has been made, continued vigilance and innovation are necessary to mitigate the impact of technology on the environment and build a more sustainable future.