It's COP30 this week. As you probably know, it’s a conference to counter global warming, to measure progress, and to set common goals.

As you also probably know, many of these goals are not met, or even flouted. Especially by the US under Trump.

But this is not about the increasingly irrelevant US. Or about how we as nations fail to meet goals. It’s about how China aims to achieve them.

But, you might say, doesn't China have the highest number of coal-fired power plants in the world, with emissions only set to increase?

Yes, but it’s also producing 60% of sustainable energy in the world. And it holds 80% of the production of solar panels, 60% of wind turbines. Coupled with hydropower and nuclear, enough for the country to reach net zero by 2060, and be world leader in sustainable energy products.

And increasingly, it’s using Artificial Intelligence to help. Here’s how.

One. To reach net zero, industry now using fossil fuels need to use electricity instead.

Steel mills in China are mostly coal-fired. But green hydrogen plants, such as the Ansteel pilot in Bayuquan, use electricity generated from sustainable to split water into hydrogen and oxygen.

Thing is, moving coal around just needs a really big fleet of trucks. But electricity is much more fickle. A fragile equilibrium between grids must be maintained, disruption may mean rolling blackouts. AI is used to help stabilise the massive energy grid that is needed to electrify the nation.

Two. Maximise efficiency. The use of electricity is not constant, there’s day and night, temperature fluctuations. Huge consumers might start their machines. AI is used to predict where power is needed most, reroute accordingly.

DeepSeek, that stunned the world in January this year for beating OpenAI’s ChatGPT, is used for such operations.

And China's Smart Grid AI. For example in manufacturing base and technology hub Hangzhou. It analyses vast amounts of data and schedules operations four times more efficiently than humans. Ensuring that the mega-city of 13 million has a stable supply during peak demand.

But isn’t using AI to efficiently use electricity self-defeating? Given it’s enormous hunger? True. That’s why new data centres, such as in Xinjiang, are driven by sustainable.

Still, while studies show that AI could save the electricity use of a country like Mexico, right now, for every unit of energy saved by AI, 1.5 units are consumed by data centre growth.

However, that's to even out as technology matures. Also, AI models are being developed for edge devices and IoT which consume far less energy.

The expectation is then, that around 2035 AI’s energy savings could exceed its consumption.

At this week’s COP30 there’s one notable cancellation. The USA. Climate change is fake, a hoax, Trump said for an incredulous UN.

But you know who’s there, and in force?

China.