Can AI help us become carbon-neutral?


Nigel Cummings

Recently (May 2019) the UN’s AI For Good Global Summit in Geneva, highlighted AI applications as possible solutions to some of the problems associated with climate change. The theme that ran through this summit was, “while most countries aren’t cutting emissions nearly fast enough. AI could help speed things up”.

The greater adoption of technology and a gradual shift to computation-heavy societies means more data needs to be crunched, and that requires energy. In the United States, data centres already account for about 2% of all electricity used. Sims Witherspoon, a program manager at DeepMind, has said that data centres and server farms worldwide, currently consume 3% of global energy. This figure will rise as consumer data generation and information needs grow.

Witherspoon thinks too, that AI could have a huge impact in other fields. “Large industrial systems consume 54% of global energy. Imagine the potential if we could apply this technology to industrial systems at large. We believe that we could affect climate change on an even grander scale”.

Reducing energy consumption is of paramount importance in reducing the effects our populations have on climate change. To help reduce this type of energy usage, technology giants responsible for huge number crunching operations have implemented AI in their data centres to optimise energy efficiency.

In 2007, Google, for example, became carbon-neutral by buying renewable energies and investing in carbon-offset programs, like planting trees and investing in wind and solar. According to Google’s senior communications officer, Ralf Bremer, achieving carbon neutrality would probably not have been possible if Google’s AI DeepMind program had not been implemented.

Since that time, the ‘Paris Agreement’ has made the world aware that we have to virtually eliminate fossil fuelled energy from all sectors of world economies. This will mean much effort is needed to network decentralised, fluctuating renewable power generation with consumers, that automatically adjusts to minimise waste energy production and help balance Earth’s overall energy requirement.

Google’s DeepMind project is certainly a ‘step in the right direction’ and may provide models that other companies could follow. It has both reduced energy waste and illustrated how AI can make a difference to climate change.

DeepMind has used years of historical data to reduce the amount of energy Google’s servers consume by controlling all of Google’s data centres cooling systems efficiently. Bremer says the DeepMind program has reduced energy consumption at its data centres by around 40% for cooling and around 15% percent in overall energy use.

The non-profit Borderstep Institute in Berlin has deployed AI to save 20% to 25% on energy used to heat a cluster of 250 apartments in the German capital. As such, this can be seen as a small step, but a step in the right direction. Others must follow.

More on the Paris agreement at: bit.ly/EnergyAI

Carbon.jpg