Build fast and write the rules

This is article 3 of 4 in What the Future Wants from Us series. Here is article 2.

For most of my working life, Asia appeared in Western sustainability conversations mainly as the scene of the crime—the coal, the cleared forest, the plastic in rivers like the Yangtze or Yellow rivers. Looking back from 2037, however, a more qualified reading turns out to have been truer.

While the West trained its attention on Asia's unsustainability, the physical infrastructure of the next economy was being designed, erected and poured in the Global East. And where the transition happens fastest, the winners often end up writing the new rules.

The future of sustainability lies in Asia

In 2025, for the first time in over half a century, coal-based electricity generation fell in China and India in the same year—not from recession, but because they added clean power faster than demand grew.

China was already making something like 80 per cent of the world's solar panels and close to 70 per cent of its EV batteries; more than half its new cars were electric or plug-in; clean technology drove close to a tenth of its GDP. Meanwhile, India hit 50 per cent non-fossil power capacity years ahead of target.

None of this was an accident. China worked systematically to seize the 'commanding heights' of the emerging green economy: the country combined early subsidies, rapid scale-up, sharp cost reduction, and control of the whole value chain from raw materials to finished product.

Today it refines most of the world's critical minerals—lithium, cobalt, graphite, rare earths—giving it strategic power far beyond manufacturing. While the West busied itself setting targets, China built the infrastructure. And, as we have seen, perfectly is happy to weaponize it.

No surprise, then, that a recent GlobeScan survey of 124 sustainability leaders for BSR found Asia-Pacific, China and Japan holding three of the top five spots for expected influence, Asia-Pacific at 62 per cent against the European Union's 51.

One caution here, however, and it matters. Building the equipment for decarbonization is not the same as delivering the full Brundtland agenda. Green Swan progress in Asia is real but incomplete. From a 2037 perspective, the physical changes may have moved fastest there, but the moral and ecological work remained unfinished. And, as we now know, the 2030s were seriously disrupted by a series of Gray and Black Swan events.

The race for smarter regulation

Looking back, the late 2020s ran three regulatory experiments at once. The United States took a chainsaw to its own rules, scrapping institutions like USAID. The EU cut more carefully, launching its Omnibus simplification in the name of competitiveness. And China quietly became an increasingly prominent rule-maker.

AI showed the three approaches at their starkest. The EU's 2024 AI Act was the world's first comprehensive, binding framework: it involved outright bans on 'unacceptable' uses, heavy obligations on high-risk systems in hiring, credit and policing, lighter transparency rules for chatbots and deepfakes, a separate regime for foundation models, and fines of up to 7 percent of global turnover.

Through the so-called "Brussels Effect", such standards tend to become global defaults—though the May 2026 Digital Omnibus delayed the toughest high-risk provisions to 2027–28, betraying growing nerves around competitiveness. The US took the near-opposite path: no comprehensive federal law, and an administration trying to pre-empt state AI laws in favour of market-led growth.

The deeper point is that markets are not God-given. They are always social constructs, and badly built ones tend to price pollution, resource depletion and social harm at close to zero. Smart regulation means moving past the crude choice between heavy control and none.

It targets outcomes rather than methods; it prices what matters, through carbon pricing and the removal of harmful subsidies; it is adaptive, using sandboxes and updatable standards for fast-moving fields; and it is durable, because no investor can plan a long-term shift against rules that may vanish at the next election.

As Gillian Tett has noted, China's subsidies are more focused and time-limited than the West's, aimed at building future industries, then withdrawn once those sectors can compete. And, typically, they do so viciously, massively accelerating the relevant cost and experience curves.

From the perspective of 2037, where the EU protected old industries, and the US often did too, China backed new ones. To the hilt. That is one of the decade's most uncomfortable findings. For much of the 2030s, the most effective industrial policy was not the most liberal one.

Why AI will be the ultimate case of both/and

By 2037, the environmental cost of AI had proved at least as bad as the skeptics had warned—in terms of land consumed, electricity burned, water evaporated by the data centers, jobs displaced, and tax revenues lost. And the same decade showed that the transition would not have happened at this speed without it. Both true at once, in short, which probably makes AI one of the purest examples we have of the potential power of both/and thinking.

The physical constraints arrived first. In 2026, AI hit what engineers began calling the "grid wall." The IEA projected data-centre electricity demand roughly doubling by 2030, to around 945 terawatt-hours—close to Japan's entire consumption—and Big Tech spent the year signing nuclear and gas deals to keep the servers fed. "Clean compute" became the phrase of the moment. And because of US trade constraints, China soon excelled in that area, too.

The money told its own story. Nvidia brushed US$5 trillion in 2026. Then DeepSeek's cheap open model wiped nearly US$600 billion off Nvidia's market valuation in a single day, while a wave of Chinese "open-weight" AI releases began turning the trillion-dollar frontier into a market commodity.

The either/or reflex splits by region, revealingly. In the US, the dominant worries are economic: job displacement, misinformation, the concentration of power in a handful of firms. In the EU, on the other hand, the anxiety is about rights, privacy, bias and democratic integrity—hence the rules-first instinct. Across much of Asia, by contrast, the mood tilts optimistic—framing AI as an engine of national advantage.

And all of this is even before the robots turn up in force. If data centers are straining modern electricity grids, millions and then billions of robots could represent a further turn of the screw.

The advisory board's take

Mei is at her most confident here, and her most disconcerting. "Whoever builds fastest writes the rules everyone else must live by. That is not a threat; it is arithmetic. You had thirty years to build this, and you just wrote reports. We built the factories. Now you would like a conversation about standards."

On the other side of the argument, Léa, again, is the most disillusioned voice. "I am the one who knows all the frameworks, which is precisely why I distrust them. I helped write rules that were rolled back the moment they bit on sunk capital."

Her fear is not that Brussels loses the standards race but that it wins a hollow victory. "We may still be a key global rule-maker in 2037," she suggests. "But a rule-maker to an economy we no longer build anything for is not a victory. We risk becoming a museum with an enforcement division!"

Then Sofía spotlights the frame both have been arguing from within. "You are discussing whose standards. I am telling you where the minerals come from. The transition arrives in my region as extraction—the same holes in the ground, the same rivers, the same people killed for standing in front of the machinery, only now the mineral is lithium, and the buyer parrots the word 'green' while signing. Write whatever standards you like. But ask who is enforcing them four thousand kilometres from your parliaments?"

Niamh, 34, grid engineer

The seventh seat now goes to someone who connects things to Ireland's electricity network for a living. Niamh has spent eleven years on transmission planning, first for a national operator, now for an independent developer. Hers is the least ideological position in the session, and by some distance the most inconvenient.

"Everybody except Mei is arguing about intent," she says. "I work on lead times. A large substation is four to seven years from consent to full production, if nobody seriously objects. High-voltage transformers run two to four years on order, from a handful of manufacturers, and the AI and renewables build-outs are now bidding for the same units. You cannot legislate a transformer into existence, and you cannot subsidize one any faster."

She is unimpressed by all three regulatory models. "The EU can pass a beautiful act. The US can repeal a beautiful act. Neither changes my queue. China builds faster than us, yes—and the reason is not mainly political will. They trained the people and built the factories that now make the equipment twenty years ago, when it was boring. Steel, cable, transformers, switchgear, and the engineers to install them. My team is short eleven engineers. We have been short roughly that number for three years."

In summary

Three voices arguing about standards, and an engineer pointing out that standards often are not the main binding constraint. Mei is right that building confers the power to set standards. Léa is right that standard-setting without production is hollow. Sofía is right that both are describing a supply chain whose furthest reaches are well beyond most people's understanding. And Niamh is right that all three assume infrastructure that does not yet exist and cannot simply be conjured by any of the three regulatory models on offer.

By 2037, however, absent a superpower confrontation of as-yet-unimagined scale, expect many standards still to be set in Brussels, while most of the technology and cost-curve progress comes from Beijing. The open question is whether Western governments can learn how to back future industries rather than old ones—effectively, in time, and at competitive cost.

This article is also published on John Elkington's Substack. illuminem Voices is a democratic space presenting the opinions of leading Sustainability Thought Leaders, their views do not necessarily represent those of illuminem.

The world needs sustainability knowledge. At illuminem, no interest group or shareholder can influence our work. Thank you for supporting our mission to make high-quality and independent sustainability information free for all. Every contribution helps. Thank you for donating today.