AI can help save a planet in peril – we have the intelligent tools to meet the challenge

Professor Letlhokwa George Mpedi is the Vice-Chancellor and Principal of the University of Johannesburg.

He recently published an opinion article that first appeared in the Daily Maverick on 23 May 2023.

Tackling the climate crisis is urgent and we need every instrument at our disposal, including artificial intelligence. Though there are valid concerns around its use, the benefits speak louder.

There are two parallel realities unfolding before us. First, we are tangibly seeing the devastating impact of climate change. Second, technological advancements in the field of artificial intelligence (AI) are gaining pace at breakneck speed without any sign of slowing down.

In 2011, in an address to the United Nations General Assembly, former UN secretary-general Ban Ki-moon cautioned: “Saving our planet, lifting people out of poverty, advancing economic growth… these are one and the same fight. We must connect the dots between climate change, water scarcity, energy shortages, global health, food security and women’s empowerment. Solutions to one problem must be solutions for all.”

Perhaps the solution lies in the convergence of AI and the fight to save our planet.

In March, the Sixth Assessment Report of the Intergovernmental Panel on Climate Change was released with startling findings. UN secretary-general António Guterres described it as “an atlas of human suffering and a damning indictment of failed climate leadership”.

As the report revealed, the impact of climate change is more severe than previously imagined and future risks are expected to escalate rapidly.Though adaptation measures – changes in processes, practices and structures  – are required, these solutions require financial backing to be scaled.

In some areas, the impact on the climate is so severe that adaptation measures cannot be implemented. Additionally, it is now imperative that the world shifts away from its overreliance on fossil fuels. The phrase “do or die” is becoming a warning that we simply must take seriously.

In Africa, this call is even more urgent as we are the most impacted by climate change. Research suggests that Africa’s climate has warmed faster than the global average, and the continent has been ravaged by cyclones, floods and extreme weather phenomena. The Climate Change Vulnerability Index for 2015 indicated that seven of the 10 countries most at risk from climate change are in Africa.

As AI gains pace alongside these shifts, hastened by the Covid-19 pandemic, there are strong arguments to be made for AI-backed solutions. According to a 2022 Boston Consulting Group (BCG)  report based on a climate AI survey, 87% of private and public-sector CEOs with decision-making power in the areas of AI and climate believe AI is an essential tool in the fight against climate change. There are a range of ways in which AI can be deployed, but as BCG notes, it cannot be used in isolation to combat the climate crisis. It is, however, a valuable tool that should be employed.

How we can use AI

AI can improve the efficiency of energy systems. Using data, AI can be used to identify areas where energy is being wasted, reduce use during peak hours, detect equipment failure, reduce overall consumption, and identify ways to reduce overall consumption.

This could be particularly effective in South Africa as we continue to face energy challenges and rolling blackouts. Moreover, the impact of climate change could be lessened by adopting cleaner power sources, such as renewable energy or smart grids, and AI-enabled electric cars and shared transport could have a similar impact.

For instance, at the beginning of this year, the University of Johannesburg (UJ) unveiled a fleet of electric buses – a testament to intelligent technologies in action. These buses have allowed UJ to reduce its carbon footprint. Additionally, UJ has made tangible shifts towards embracing renewable energy in recent years, and solar energy now makes up 15% of its energy sources. This is a reminder that universities have a responsibility to provide solutions to our most deep-seated challenges.

In 2015, UJ established the Institute for Nanotechnology and Water to realise the potential of nanotechnology applications to alleviate many of South Africa’s water problems, the results of which can be applied to other countries in Africa.

Elsewhere, AI can help us to better understand and predict the impacts of climate change. By analysing large amounts of data on weather patterns, sea levels and other environmental factors, AI can model the effects of climate change on ecosystems, human health and infrastructure. This can also be used to identify strategies for adapting to these changes.

Microsoft, for example, has built a planetary computer designed to work as a geospatial search engine to expedite climate decision-making and avert the impact of natural disasters.

While there are certainly impactful AI solutions that can be implemented, this shift is not without its challenges. For example, AI in itself can be energy-intensive, particularly when training large-scale models. This raises concerns about the carbon footprint associated with AI technologies and whether this is an effective trade-off.

There are also various ethical considerations. AI systems have demonstrated bias, and in the deployment of AI solutions there needs to be an emphasis on fairness and the equitable distribution of resources and benefits.

Then there are implementation and policy challenges. The regulations surrounding AI are pretty scant. Widescale implementation requires new policy frameworks, supportive regulatory frameworks and the buy-in of stakeholders.

The data limitations, particularly in certain regions, also need to be considered. Data gaps, inconsistencies and limited access to relevant data can hinder the effectiveness of AI models and limit their ability to provide accurate predictions and recommendations. In tracking the progress of the UN Sustainable Development Goals, for example, it has been challenging to find similar data sets across regions.

Though these concerns are valid, the argument for AI solutions seems to outweigh the challenges. If we are indeed in the age of AI, it stands to reason that sustainability and advancement should be seen as mutually inclusive concepts. One simply cannot be achieved without the other.

*The views expressed in this article are that of the author/s and do not necessarily reflect that of the University of Johannesburg.

Share this