Over the past couple of years, the volume has been turned down on ESG (environmental, social and governance) – the claims and the rhetoric, that is. As for the actions and the results, they have continued to forge ahead regardless.
I think this has been a helpful reset for sustainability, one that’s more honest and more focused on what can be achieved. For a while there, sustainability became too anchored in aspiration. There was too much talk about what was going to happen down the track, and not enough focus on what companies were actually doing today. It often led to wildly optimistic forecasts and subsequent scepticism that goals could be achieved.
What makes me excited about the next five years for sustainability is that while we’re hearing about it less, the improvements are still there. I think it’s because of a solid truth at the core of sustainability: less waste is cheaper than more waste, happier workers are more productive, and cleaner production avoids the higher costs of pollution down the road. This stuff makes sense – and that's why companies keep doing it.
Old king coal’s unsustainable restoration
Take carbon emissions from energy consumption in the US (mostly transportation and power generation): in 2024 they were at their lowest level since 1987 (excluding the pandemic). Yet the economy is 2.5 times larger, and supports 40% more people. Per capita, it’s the lowest since 1940, which is mind-blowing when you think that only half of Americans had a single car back then. Today the average family has roughly two. And 90% of American households now have air-con – in 1940 it was virtually nonexistent in homes.
Electricity’s moment
The next five years are going to be very exciting for electricity. For the past quarter century, electricity generation growth in the Western world has been very low – the G7 has averaged just 0.3% a year, according to Enerdata. Meanwhile, the demand side picked up the slack by becoming phenomenally more efficient. We’re using more gadgets each day and producing more than ever, and there are billions more of us.
Electricity now accounts for a quarter of global energy use, up from 15.5% in 2000. That trend is likely to accelerate. After decades of underinvestment in generation and the network infrastructure that gets power where it’s used, the investment is flowing. There’s a lot of concern and comment about how much power AI is using. Yet, while the expansion is tremendous, it’s still nowhere near the largest demand driver over the next five years (it’s a big portion of the data centre growth in the chart below – but not all of it).
After decades of being treated as a mature, slow-growth industry, electricity is quickly becoming one of the most strategically important commodities in the world. The reason is AI. Global data centre electricity demand grew 17% in 2025, with AI-focused data centres growing a remarkable 50% – well outpacing overall electricity demand growth of 3%. The International Energy Agency expects total data centre electricity use to roughly double by 2030, reaching close to a trillion kilowatt-hours. While that’s roughly the same amount of power that Japan uses today, it would still account for just 3% of global electricity use.
Yet the technology is often heavily clustered: research by smartgrid expert Schneider Electric (which we own) suggests AI alone could account for up to half of all US electricity demand growth between 2025 and 2030.
That's a staggering demand shock. And it's coming at exactly the moment when grids globally need significant investment after decades of underspending. The encouraging part is that AI itself is also driving rapid efficiency gains – and commissioning huge amounts of new power. Power consumption per AI task is falling by roughly an order of magnitude every year. Similar productivity gains are showing up across science, technology, manufacturing and clean energy generation itself.
Nationalism, scarce resources, and the case for "made nearer home"
Another, less benign, theme for the rest of the decade is the rise of economic nationalism. In some ways, this is a healthy acknowledgement that there’s a trade-off between efficiency and resilience. But it can spiral into a zero-sum framework that makes everyone worse off and crimps global development.
Countries are increasingly trying to secure a reliable supply of power, materials, chips and other critical inputs while often putting barriers in the way of geopolitical opponents (and sometimes even friends).
We hope this moderates and drives greater focus on efficiencies at home, rather than conflicts abroad. Finding ways to recycle the resources we have, rather than throwing them out and getting cheaper fresh materials from overseas could be a useful spur for doing more with less (productivity by another name). It could also focus our minds on what we use to begin with – preventing waste in the first instance.
This economic nationalism could be a reaction to increased scarcity of resources, whether power, fertiliser, water, rare minerals, or myriad other things crucial to modern life. We own several businesses that are trying to make these bedrock needs more efficient and better yielding. One of them is tractor maker Deere. It’s more than just powerful machines that help farmers work their farms. Deere realised long ago that GPS technology, combined with cloud computing and sensors, can help farmers keep track of every square metre of their sprawling lands. Monitoring for exactly how much water is required, keeping nutrient levels balanced without over-fertilising, and even the most efficient way for harvesters to make their way around the fields.
Deere's See & Spray technology can cut herbicide use by up to half while maintaining yields – better economics for the farmer, less chemical run-off into rivers. Precision agriculture is targeting nutrients and water far more accurately, reducing waste in one of the most resource-intensive sectors on the planet. This is sustainability that pays for itself.
Cleaning up the not-yet-clean
These are the sorts of uses for AI that could spill out into other industries as well. We don't see that reversing, which is why we’re optimistic about the potential for AI to stimulate and improve industries across the economy. That’s why we’re investing in the businesses that make the chips, develop the tools, and build the infrastructure to make our power networks greater and more efficient.
One point that often gets missed: many of the same technologies driving the clean-tech opportunity are also making dirtier, more carbon-heavy industries materially cleaner. Better sensors, software and equipment are helping water companies find and fix leaks before they waste millions of litres. Mining companies are using digital and chemical innovations to get more metal out of the ground with less pollution and better yields. Industrial users are cutting water and chemical use in production. If we can support the businesses making tech that helps clean up the dirtiest industries – which often produce raw materials that are critical for most modern products – that represents enormous, real-world progress.
A more cheerful conclusion than you might expect
A lot of people seem gloomy about the state of the world, and gloomy about the state of sustainable business practices. I get it. But the reality is that the world – while clearly not perfect – gives more people a better life today than has ever been the case in human history.
Sustainable investing has not been an easy place to be for the last five years, yet it’s still delivered substantial progress. And the next five are shaping up to be very exciting indeed: the investment needed in grids and networks is finally happening, the efficiency gains being unlocked by AI are genuinely remarkable, and the companies quietly doing the right thing tend to be the ones delivering for shareholders too.
We think that's a story worth being invested in.
