On AI/ML and the Environment
On June 10, the report “Tackling Climate Change with Machine Learning” was released. Contributors included researchers at University of Pennsylvania, Carnegie Mellon, University of Colorado - Boulder, Harvard, MIT, Cornell, Stanford, DeepMind, Google AI and Microsoft Research. The following are selected comments from the report:
“Machine learning, like all technology, does not always make the world a better place – but it can … ML can make systems more efficient (e.g. prevent electricity loss during transmission, consolidate freight, and reduce food waste). It can enable remote sensing and automatic monitoring (e.g. pinpoint deforestation, gather data on buildings, and track personal energy use). ML can provide fast approximations to time intensive simulations (e.g. climate models and energy scheduling models), and it has the potential to lead to interpretable or causal models (e.g. for understanding weather patterns, informing policy makers, and planning for disasters)."
“Electricity systems are currently responsible for about a quarter of human-caused greenhouse gas emissions … To reduce the impact of electricity systems across the globe, society must: transition to low-carbon electricity sources (such as solar, wind, hydro, and nuclear) and phase out carbon-emitting sources (such as coal, natural gas, and other fossil fuels), reduce emissions from existing carbon-emitting power plants, since the transition to low-carbon fuels will not happen overnight.”
"The transportation sector accounts for about a quarter of global energy-related CO2 emissions … Strategies to reduce green house gas (GHG) emissions from transportation include: decreasing transportation activity, increasing vehicle efficiency, reducing the carbon impact of fuel, shifting to lower-carbon options, like rail."
Buildings offer some of the lowest-hanging fruit when it comes to reducing GHG emissions … It is possible today for buildings to consume almost no energy … Machine learning provides critical tools both for managing buildings and for designing policies surrounding them."
Plants and algae have been accumulating and sequestering carbon through photosynthesis for millions of years … Our current economy encourages practices that are freeing large amounts of this sequestered carbon through deforestation and unsustainable agriculture … The large scale of this problem allows for a similar scale of positive impact, and ML will play an important role in many of these solutions."
AI/ML technologies may not result in “silver bullet” environmental solutions, but they may lead to insights that drive better decisions.
AI/ML will require many other technologies to help 1) monitor the environment; 2) identify inefficiencies in emission-heavy industries and 3) model complex systems which can be used to adapt to environmental changes.
Better data and better models can help develop potential solutions – addressing the problem will require behavior modification by individuals, businesses and governments around the world.