Eric Schmidt

Superpower truce buckles under Trump election meddling claims Geopolitical friction between Washington and Beijing has flared once again, threatening the fragile diplomatic stabilization established earlier this year. In a series of public statements, former U.S. President Donald Trump alleged that China executed a massive influence campaign targeting over 200 million American voter files during the 2020 U.S. presidential election. Beijing immediately struck back. The Chinese foreign ministry dismissed these claims as entirely fabricated political theater designed to vilify the country. This dispute threatens the constructive relationship framework established during the bilateral summit in May. At that meeting, Trump and Chinese leader Xi Jinping agreed to build a framework of strategic stability. Analysts on the podcast China Decode, hosted by Alice Han and James Kynge, suggest the timing of Trump's allegations is highly calculated. The claims emerge just two months before Xi is scheduled to visit the United States, raising concerns of a premature end to the superpower truce. However, the quiet response from the current White House indicates these claims may serve domestic political goals rather than representing a shift in official state policy. Trump remains focused on relitigating his 2020 electoral loss while building pressure ahead of upcoming domestic elections. Significantly, Xi's invitation to visit Washington on September 24th remains active, and Trump is still considering attending the APEC summit in Shenzhen this November. This diplomatic continuity suggests that while public rhetoric remains hot, the underlying bilateral channels remain open. Global opinion registers a startling tilt away from Washington While Washington politicians trade barbs, the broader geopolitical theater is shifting in Beijing's favor. Recent data from the Pew Research Center reveals a stunning decline in international confidence toward the United States relative to China. Currently, 46% of people surveyed globally view China more favorably than the United States. This represents a massive reversal from 2023, when Pew data showed the U.S. holding a comfortable 58% favorability rating compared to China's 32%. This shift is most acute among America's traditional European allies and immediate neighbors. In the United Kingdom, Germany, and France, public sentiment now favors China over the United States. This trend persists despite the UK's deep structural and military dependence on the U.S.-led NATO alliance, and Germany's ongoing industrial competition with Chinese manufacturing. Public opinion in Mexico shifted by 19 percentage points in favor of China, while Canada registered an 11-point swing. This realignment points to a systemic failure in the Western Hemispheric approach, which was heavily emphasized in Washington's recent defense strategies. The primary driver of this shift appears to be Trump himself. European populations find Trump's erratic policy positions deeply unsettling. In contrast, Beijing has projected a message of relative policy stability and successfully executed a global soft power campaign, bankrolling massive infrastructure investments throughout the developing world. Capital efficiency and the open-source offensive In the technology sector, the battle for dominance in artificial intelligence is moving at an extraordinary pace. Selina Xu, head of China research for former Google CEO Eric Schmidt, notes that technological leadership in large language models has become a highly perishable commodity. Startups on both sides of the Pacific hold performance leads for a month at most before a rival system eclipses them. Chinese developers are displaying remarkable motivation and clarity of purpose. Following a period of introspection after the release of DeepSeek R1, Chinese founders are executing a highly coordinated strategy. They are aiming directly at American leaders like Anthropic, using coding agents as their primary development target. China's AI ecosystem is building this momentum on two distinct structural pillars: * **Commitment to Open Source:** Unlike closed American labs, Chinese giants and startups are keeping their models open. This strategy was reinforced by Xi Jinping at the World AI Conference. * **Compute and Capital Efficiency:** U.S. chip export controls have backfired in a crucial way. By restricting access to state-of-the-art silicon, Washington forced Chinese firms to pioneer extreme optimization techniques, extracting maximum training efficiency from limited hardware. This push culminated in the release of Kimi K3 by Moonshot AI, a massive model sporting 2.8 trillion parameters. Kimi K3 demonstrates that Chinese open-source systems are no longer just cheap, lightweight alternatives; they are competing directly at the frontier of raw intelligence. Beijing soft-pedals the algorithmic companion While American AI discourse is often dominated by Silicon Valley's theoretical debates over Artificial General Intelligence (AGI) and existential risk, China's relationship with AI is deeply pragmatic. This pragmatism is shaped by a unique feedback loop between private tech entrepreneurs and the state. Following the regulatory crackdowns of the pandemic era, Beijing has realized it needs the private sector to drive economic growth, yet the state remains hyper-vigilant regarding social stability. This delicate balance is highly visible in how Beijing regulates AI companions and autonomous systems. The government is acutely aware that the gig economy employs roughly 320 million people—representing 44% of the nation's workforce. To prevent mass labor disruption, the state has intentionally slow-rolled licensing for autonomous vehicles. Similarly, regulations targeting AI companions reflect deep worries over societal stability and demographics. Fearing that young citizens might substitute real romantic relationships with customized AI bots, thereby exacerbating China's declining fertility rate, the government issued strict draft guidelines in December. However, showing the collaborative nature of Chinese tech policy, the final regulations were heavily watered down following industry feedback, removing mandatory 24/7 human-takeover requirements and simplifying data training rules. This compromise protects the commercial viability of startups while preserving state oversight. The search for the robotic brain On the hardware front, China's manufacturing base gives it a massive advantage in the humanoid robotics sector. Companies like Unitree, Xiaomi, and UBTECH are producing highly advanced physical platforms. However, both Chinese and American developers face a critical bottleneck: the creation of a generalized robotic brain. Currently, humanoid robots are deployed in highly controlled, specific environments, such as performing simple, repetitive tasks on electric vehicle factory floors. These deployments are primarily used to collect physical interaction data. When placed in the unpredictable, irregular settings of a real-world home, even the most advanced robots fail. They lack the physical reasoning required to adjust to minor changes, such as a cup resting at an unusual angle. To bridge this gap and achieve a breakthrough comparable to the chatGPT moment in language modeling, researchers are divided. Some favor scaling multimodal language models to handle physical tasks. Others argue for the development of dedicated "world models" that natively understand physical laws. To accelerate this transition, Chinese firms are building dedicated robotic data factories, using physical simulations alongside real-world testing to train their machines at scale. Semiconductor export machine doubles its global footprint Looking to the horizon, the intersection of macroeconomics and hardware manufacturing will define the next phase of global trade. Despite extensive Western efforts to ring-fence China's tech sector, Beijing's export machine is showing incredible resilience. In the first half of this year, Chinese semiconductor exports to the global market doubled in value compared to the previous year. This massive surge is driven primarily by memory and DRAM chips, where Chinese suppliers like CXMT are injecting vast amounts of low-cost supply into global supply chains, directly challenging established South Korean giants. This commercial success stands in stark contrast to China's sluggish domestic consumer market, where retail sales climbed a meager 0.3% in the first half of the year. Kynge predicts that full-year retail sales growth will remain below 3%, indicating that while China's domestic consumer recovery remains muted, its high-tech export engine is running at full throttle.