Environmental impact of data
A google search, a scan of your passport, CCTV footage, Ai tech, an Instagram upload, an email or streamed video. These entities generate information about our daily lives, known as data.
Transferring data requires the collaboration of computers, phones, under water powerlines, satellites and many data centres. Collectively these attributes form data ecosystem. The industry emits 3.7% of
global emissions, rivalling the air industry (energuide.be, 2021).
With a growing number of 4.66 billion internet users worldwide, our digital consumption is no less problematic than material goods (Johnson, 2017).
24 hour data farms used to transmit, store and process data require vast amount of energy, water and land. Their locations are often biased and can have damaging effects on social welfare and landscapes (Lucivero, 2019).
Metaphoric terminology used to encourage end users, neglects to inform people that the digital space has a physical body.
Data intensive systems such as AI are often presented as more eco-friendly solutions to managing health, weather, businesses, farming and shopping. Even though the technology has potential for positive change, the large complex amounts of algorithmic data needed to run AI is extremely uneconomical.
A graphic card required to run the average ai software uses as much CO2 as a flight across America (Atmos, 2021). Training a complex AI uses 2.8 gigawatt of electricity per hour (Knight, 2020) and emits 5xs more CO2 than a car (Atmos, 2021).
Governments have the power to enforce corporations to change the data terminology, user consumption habits, encourage more sustainable methods of data processing and collection.
Institutional policy change will only be recognised through public pressure.
Questioning the individuals daily data usage, will start a conversation and interest on this issue, leading to action. Public demand for policy change will help the environment as data
companies will be forced to make sustainable and socially responsible choices when implementing data technologies.
