Opinion | China Bets on 3-Cent Green Power to Win the AI Race

2026-08-21 15:00
Since 2007, the Summer Davos Forum has been held annually in China, rotating between the cities of Tianjin and Dalian. (Photo by Jason Wang)
Since 2007, the Summer Davos Forum has been held annually in China, rotating between the cities of Tianjin and Dalian. (Photo by Jason Wang)

At this year's Summer Davos Forum in Dalian, something unexpected happened. China's hottest sectors — robotics, autonomous vehicles, large language model labs — were largely absent from the main stage. The spotlight belonged instead to a quieter group: the engineers and executives who build the power infrastructure underpinning all of it.

The message they delivered, across three days of panels and sideline conversations, was pointed: green electricity is China's primary weapon in the competition for AI dominance. Not chips. Not algorithms. Power.

Energy Leaders Upstage AI Giants at Davos in Dalian

The lineup at New Champions 2026 — the formal name for the Summer Davos gathering, held in Dalian in late June — read like a who's who of China's energy transition. CATL founder Zeng Yuqun outlined his plan to turn China's 40 million electric vehicles into a distributed power and computing network. LONGi Green Energy chairman Zhong Baoshen sketched a future in which solar generation and storage merge into a seamless baseload system. Wu Zuyu, chairman of HiTHIUM — China's second-largest energy storage battery maker — made the boldest forecast of all.

Shanghai startup Matrix Origin deployed its Matrix-3 robot at the venue for a coffee performance at 8:30 a.m., though it spent most of its time idle. (Photo by Jason Wang)
Shanghai startup Matrix Origin deployed its Matrix-3 robot at the venue for a coffee performance at 8:30 a.m., though it spent most of its time idle. (Photo by Jason Wang)

Within five years, Wu told the audience, China would enter what he called the "3-jiao era" of green electricity. At that price — roughly 0.30 yuan per kilowatt-hour, or about NT$1.42 — clean power would cost less than coal, which currently trades at around 0.362 yuan per kilowatt-hour. The cost breakdown Wu described: one jiao to generate, one to transmit, one to store, with each component driven down by advances in sodium-ion battery technology. Unlike lithium, sodium is abundant and cheap; HiTHIUM projects sodium batteries will enter commercial production within five years, with charge cycles reaching 20,000 — a lifespan that makes per-kilowatt-hour storage costs negligible.

State-owned enterprises joined the chorus. State Grid Corporation CEO Zhang Wenfeng outlined an AI-driven "compute-electricity coordination" system capable of forecasting solar and wind output to the microsecond. Sinopec Chairman Zhao Dong described plans to convert petrol stations into green hydrogen hubs. Even the chairman of Ningxia Baofeng Group — a coal company — announced a pivot to green hydrogen, pledging 5% annual output growth.

Hydrogen-powered vehicles, symbolizing the future of new energy, have taken on the role of chauffeuring VIP guests at Summer Davos. (Photo by Jason Wang)
Hydrogen-powered vehicles, symbolizing the future of new energy, have taken on the role of chauffeuring VIP guests at Summer Davos. (Photo by Jason Wang)

China's Green Power Could Undercut Coal Within Five Years

The ambition is grounded in real momentum. China's wind and solar generation has grown 400% over the past eight years. Renewables now account for 38.5% of total power output, up from 26.7% in 2018, and are on track to exceed 40% in 2026. Renewable capacity already surpasses 60% of China's total installed electricity capacity.

The policy scaffolding arrived the day after Summer Davos closed. Beijing released its 15th Five-Year Plan for energy systems, committing 20 trillion yuan to green infrastructure with a target of pushing wind and solar above 50% of total generation by 2030. Coal — long the "ballast stone" of China's energy security — would be demoted to a backup role, dispatched only to stabilize the grid when renewable output fluctuates.

The AI imperative is explicit. China's National Energy Administration chief Wang Hongzhi noted that generating five seconds of AI video consumes as much electricity as charging 10 smartphones. Aggregated across the economy, AI demand is projected to add 600 billion kilowatt-hours to China's annual electricity consumption over the next five years — roughly twice Taiwan's entire annual power usage of approximately 290 billion kilowatt-hours. Liu Shijin, an adviser to the China Council for International Cooperation on Environment and Development (CCICED) and a senior government policy architect, put the strategic logic plainly at the forum: "Green electricity is the foundation of AI development — this path has been opened."

If the 3-jiao scenario materializes, the competitive implications for AI are severe. DeepSeek's V4 model already charges one-tenth of Anthropic's comparable pricing for equivalent tasks; ultra-cheap domestic power would widen that gap further.

Tufts and Johns Hopkins Scholars Warn of Trade Blowback

Not everyone at Davos was persuaded this ends well for global markets. Kelly Sims Gallagher, dean of the Fletcher School at Tufts University, argued that China's near-monopoly across clean energy supply chains — solar panels, battery storage, EV components — will not go unchallenged. "Countries like India and Brazil," she said, "will not want this kind of monopoly to continue" and will likely respond with trade restrictions, tariffs, or localization mandates. Green energy products risk being repurposed as geopolitical tools.

Yuen Yuen Ang, a professor of political economy at Johns Hopkins University, identified the underlying structural tension. If Chinese domestic demand cannot absorb the capacity its green energy build-out creates, the surplus will hit global markets as a new wave of disruptive exports — what she called a continuing "shock" to the world economy. Premier Li Qiang's framing of the green transition as "China Opportunity 2.0" depends on a consumption boom that China has so far failed to generate. With per capita income approaching $14,000 — a level at which consumer spending historically accelerates — Chinese households have not followed the expected pattern. Annual births have collapsed from roughly 20 million to fewer than 8 million, compressing the domestic consumer base structurally.

Those mismatches make continued export pressure a near-certainty, whatever Beijing intends.

Energy Self-Sufficiency and the Risk to Taiwan's Security

For Taiwan, the stakes extend beyond trade competitiveness. The island currently pays roughly NT$5.6 to NT$6 per kilowatt-hour for green electricity — approximately four times the cost HiTHIUM projects for China within five years. That gap, if it closes on China's terms, creates a durable cost advantage in AI and manufacturing that no amount of semiconductor export controls can fully neutralize.

The deeper question is strategic. Hong Kong investor Wang Junru, who participates in cross-border carbon credit markets, offered the bluntest framing at the Dalian sidelines: "In the current geopolitical environment, green electricity is China's most certain, least-interfered-with, most promising strategic bet — one that enables AI deployment and economic growth simultaneously."

A China that runs its AI economy on cheap, domestically sourced green power is a China with fewer energy vulnerabilities to exploit — and greater confidence for long-term strategic competition. Whether that translates into greater willingness to act across the Taiwan Strait, or simply into a widening economic and technological gap, is a question Taipei cannot afford to treat as rhetorical. (Related: Opinion | At Davos, China's Li Qiang Tears Down Walls. For Now. Latest


You've read it. Now join the conversation — follow us on X,  Facebook and IG. Editor: Penny Wang


Latest
The Real Threat to Taiwan's Motorcycle Industry Is Complacency
Beijing Watch | Xi Jinping Leads State Farewell For Reformist Premier Zhu Rongji
Beijing Watch | Chinese Podcaster Who Charmed Nolan Faces Plagiarism Allegations
US Treasury Doubles Bond Buybacks as Bessent Battles Rising Yields
Cohesity Expands Aspire Global Partner Program with New Specializations
Opinion | Why CATL Wants to Be the Infrastructure of the AI Age
Opinion | Taiwan's Trade Pivot Trades One Dependency for Another
TPP Support Drops to Historic Low in Latest Taiwan Poll
IGEL Brings Now & Next Workspace & Endpoint Security Summit to Melbourne, Australia
Sands China Wins Two PATA Gold Awards
Chongqing Chaos: TFBOYS Agency Trainees Targeted by Online-Organized "Anti-Fan" Mob in China
Binance Introduces Agent OS to Connect AI Applications to Financial Infrastructure
Beijing Watch | China's Growth Slows to 4.3%, Consumers Stay Home
Malaysia Warms to China, Cools on America: Reading Anwar's Pivot
Dyna.Ai, DOLBIX Consulting, and Nikkoku Soft Launch AI-Powered Insurance Sales Support Pilot
Life Sentence for Evergrande Founder Hui Ka-yan as Fraud Case Concludes
China's 'Blue Whale' Sub Drone Nears Deployment Amid Dual-Use Concerns
DB HiTek to Make Debut at electronica India 2026, Targeting Growth in Indian Market
Opinion | Beyond the Seabed: Building Taiwan's Satellite Resilience
"Spicy Happiness" with aespa Nongshim Launches Second Global Shin Ramyun Campaign