Editor's note: Hua Gesheng is a commentator on international and multilateral affairs. The article reflects the author's opinions and not necessarily the views of CGTN.

Human civilization has always progressed on the strength of great technological leaps. Yet on the eve of every such leap, hope and unease have gone hand in hand. History shows that what ultimately decides the fate of any technological revolution is not its sheer power, but people's capability to harness its benefits, guard against its harms, and make it fit for purpose.

Today, humanity stands on the threshold of yet another transformative moment. Over the past year, artificial intelligence (AI) has advanced by leaps and bounds. On some of the most demanding benchmarks ever devised, top-performing AI models have reached parity with human performance. The policy decisions we make today will determine whether we share in the rewards of a productivity boom or bear the costs of a frayed social fabric – and whether the next generation of technology will be clean, open, and people-centered, or costly, divisive, and misaligned with human purposes.  In this regard, China has a unique solution to offer.

The moonshot moment

During the 2026 World AI Conference (WAIC), Chinese AI start-up Moonshot unveiled Kimi K3, the largest open-weight large language model that performs neck-and-neck with some of the most powerful proprietary models from Anthropic and OpenAI. Commentators describe this as "the moonshot moment": The first time a fully open-weight model has been that close to the actual top of the leaderboard, rather than trailing a tier behind.

While countries are concerned with the US administration's recent decision to shut down access to a frontier AI model, Chinese open-source and open-weight models are gaining global traction. When AI models are freely shared, developers, researchers, and entrepreneurs everywhere can adopt and build upon them, thus accelerating innovation and distributing benefits more broadly. If access to cutting-edge AI remains concentrated among wealthy nations and a handful of dominant firms, existing inequalities are likely to deepen: Countries and communities already lagging in digital infrastructure will fall further behind, widening the digital and wealth divide globally.

Openness is not merely a business choice of Chinese tech companies. It has also become a hallmark of China's national AI strategy. Regarding openness as a vital pathway to inclusive AI development, China stays committed to fostering open-source ecosystems, promoting the exchange of AI research outcomes and technical know-how, and enhancing the accessibility of AI technologies and services. This is exemplified by China's announcement at WAIC that it would provide developing countries with 5,000 AI training opportunities and make its AI-powered meteorological warning system MAZU available to 30 countries.

It's wise cost-wise

As usage has grown, so have token volumes – and with them, the bills. Even with open-weight models, many countries remain deterred by the high costs of AI infrastructure and its enormous energy demands. What was once a story about a productivity leap is turning into a reckoning: Is AI too costly to be useful?

Chinese labs have turned this constraint into an advantage. Forced to operate with downgraded hardware as a result of US export controls on chips, Chinese developers have learned to extract far more value from each processor through algorithmic efficiency. Start-ups like DeepSeek and MiniMax have trained their models at a fraction of the expenditures by other frontier labs, driving down computing costs by maximizing the utility of every token.

Affordability is only part of the equation; environmental and energy costs also matter. The massive data centers powering large-scale AI systems consume vast amounts of electricity and water, straining local grids and water supplies. Yet by contrast, energy costs in data center regions in China are 30-50% lower than in the US, thanks to advantages in green electricity. The efficiency of Chinese models and the country's energy sector offer a compelling alternative: a path toward AI development that is economically more affordable and environmentally more sustainable.

A tool that truly empowers

AI has dominated public discourse for much of 2026, yet responses to its adoption have varied widely across the world. In many countries, workers fear that AI threatens their livelihoods rather than enhancing them – concerned that its gains are being captured disproportionately by a small group of tech firms, chipmakers, and their investors. At the same time, the costs and disruptions fall broadly on workers and communities who have little to gain in the technology's rollout.

To avoid this scenario, China has introduced several concrete measures to harness AI towards a pro-job direction. Its "AI Plus" guidelines call for broader AI applications that create new jobs and upgrade existing ones, while steering innovation resources toward sectors with strong employment potential. The guidelines also promote "human-machine collaboration" and encourage the deployment of AI in dangerous or understaffed industries – pivoting the country from leveraging a "demographic dividend" to cultivating a "talent dividend."

The ancient Chinese philosopher Zhuangzi held that people should "use things, and not be used by them." His wisdom offers a timeless framework for thinking about technology: We must amplify its benefits while sidestepping its pitfalls. The same principle underpins China's approach to AI development and governance. AI should be harnessed to empower humanity, unlock new opportunities, and improve well-being – while AI governance should ensure that its risks are minimized and its benefits widely shared. This is how technology can truly fulfill its mission: as a tool that genuinely empowers.

(If you want to contribute and have specific expertise, please contact us at opinions@cgtn.com. Follow @thouse_opinions on X to discover the latest commentaries in the CGTN Opinion Section.)