Another week, another warning from eminent scientists that climate inaction is set to cost us and our planet dearly. A survey of hundreds of members of the Intergovernmental Panel on Climate Change (IPCC) published on May 8th, shows that 8 in 10 of the scientist respondents expect global warming to hit at least 2.5ºC above pre-industrial levels this century.
While some may read this and feel like the only option is to curl up into a ball of despair, I want to make the case for hope.
I believe that by mindfully creating the right technologies and using them well, we humans do yet have a chance of bringing about serious change. I’m not talking about silver bullets. When I say the “right” technologies, I mean those which allow us to see the true picture clearly, communicate well, and free up time for us and our companies and governments to take action, rather than simply tallying up climate change data.
As co-founder of a sustainability data management software platform, I get up close and personal with these topics on a daily basis, and want to share with you my specific optimism for the power of Artificial Intelligence to play a more significant role in this story than has been generally predicted up to now.
The first thing I ought to say is that I am not prone to hype. I am Swiss, very pragmatic, maybe even a little boring. At my company, Sweep, we didn’t get carried away with Web3.0 and NFTs. But I am all in for the already-much-hyped AI, and here’s why.
The company I co-founded, Sweep, is all about something that most people find very uninteresting: data. But interesting or not, data is all around us, it’s constantly being generated by us and the companies and organisations we work for and interact with, and when you start collating and analysing it, you give yourselves the keys to making things better. Our platform helps organisations specifically to act on carbon and ESG data – from mapping to measuring, reducing to reporting. Once you have the data building blocks, you can scale your climate action across your teams, supply chains, and portfolios quickly and cost-effectively.
I have to admit that one problem with data is that while it is everywhere, it’s usually not in a handy, uniform format. The bigger and more complex your organisation, the more likely it is that this issue will be magnified and the time you will spend trying to wrangle your data into shape is multiplied. Then when you start working towards the different compliance requirements and frameworks, you need to present different bits of your original data in different ways. That meme of the lady looking confused, with mathematical symbols appearing over her face, would feel appropriate here.
Artificial Intelligence is not a definitive solution to these data issues in and of itself. You can’t just plug some AI into your hard drive and type “create a carbon reduction plan” and let it go. Maybe one day, but not quite yet.
However, when you weave Artificial Intelligence into your sustainability management practices, you can cut through mountains of the drudge work. I’m talking about data-gathering, formatting and error-fixing. And writing pages of reports that reference and compare your data. The really dull and time consuming stuff. You get to carve out valuable time for the important parts: Analysing the data you and your AI have collated, making data-based decisions and implementing the change; Uncovering hidden opportunities for emissions reductions and for lessening other negative impacts on nature; Taking action to reduce your climate impacts wherever and however you can; Setting realistic goals based on meaningful figures; Pinpointing links in your supply or value chain where big sustainability wins can be made.
And here’s the bit where the money people rightly get excited: figuring out how to gain business advantage through adapting your company to the low-carbon economy of the future.
Of course, knowing your own data is just one part of the story. It’s also important for any organisation to be able to assess and manage its climate-related risks. It’s a complex task, but by leveraging the power and speed of AI, vast amounts of data can be analysed to develop sophisticated risk models.
One great example is the groundbreaking “Project Gaia”, developed by the Bank for International Settlements, with its partners the German Bundesbank, the Bank of Spain, and the European Central Bank. This project has already produced tangible results by using AI to survey and identify climate-related risks in the financial system.
Personally I would argue that when the people holding the purse-strings of our international financial system are taking climate related risks so seriously, it’s a good sign for the planet. It’s also a stepping stone to further AI-powered climate advances elsewhere, as this project was intentionally designed to be a model for AI-enabled applications in a broader range of use cases for central banks and the private sector.
It’s important to acknowledge that Artificial Intelligence powered technologies do come with their own environmental cost.
Companies need to be transparent about their carbon footprint and work on reducing it, whether that’s manufacturing a product or using AI or cloud hosting. To keep the climate benefits outweighing the climate harm, there are some key considerations:
First, remember that AI is just one tool in the toolkit. Know that smaller models trained for specific tasks tend to use less energy than large general use Large Language Models (LLMs). Calculate how much energy the AI uses, versus its speed, performance and accuracy. Consider whether this is based on actual information rather than estimates – and work to eliminate the estimates wherever possible. And be mindful of the energy use linked to training your AI as well as its real implementation.
To circle back to where we started – of course we must not lose sight of the fact that we have already damaged our planet, and that if we continue as things currently are, then we are not on a good path. But that is still a big if. There’s nothing inevitable about what happens next, and we do have choices: political choices, strategic choices as businesses and financial institutions, choices as consumers of who we give our custom to, and a growing range of technological choices. Artificial Intelligence on its own is not going to “save” us. But if we choose to, we can grasp the incredible possibilities that these technologies present.
