Xi pushes global AI cooperation in Shanghai. Valley campuses eye chip curbs
At Shanghai's World AI Conference, Xi urged global AI rules and criticized U.S. tech limits. Fresno State and UC Merced track chip policy that can raise lab costs.
Xi pushes global AI cooperation in Shanghai. Valley campuses eye chip curbs
Key Takeaways
- Xi Jinping urged global cooperation on AI at Shanghai’s World AI Conference and criticized U.S. tech restrictions.
- U.S. rules have repeatedly limited advanced AI chip sales to China, including Nvidia’s A800/H800-class parts.
- Fresno State and UC Merced run AI efforts tied to agriculture and teaching that rely on GPUs and cloud services.
- Any tightening of export controls could raise hardware or cloud costs for Valley labs and startups.
He said the world should treat artificial intelligence as a shared project. Chinese President Xi Jinping also called U.S. technology curbs an "overstretching" of national security, speaking Friday at the World AI Conference in Shanghai. For labs at Fresno State’s Center for Irrigation Technology in Fresno and at UC Merced, the fight over chips and rules shows up in prices and delivery times.
The chips are the tell.
What Xi said in Shanghai
Xi framed AI development and governance as work that should be done with, and not against, other countries, and he warned against politicizing technology. United Nations Secretary-General António Guterres addressed the same meeting. China’s current five-year plan through 2030 prioritizes AI and other research areas, and the remarks come as Beijing seeks more partners to keep building systems at scale.
Why this matters in the Valley
Most Valley research groups do their AI training and testing on U.S. cloud services or on campus hardware, but both track the cost of access closely. Fresno State has stood up an AI campus portal and training effort, while its irrigation center lists an AI model for estimating crop water demand among current projects. UC Merced runs ag-tech and AI irrigation research through CITRIS, AgAID and related programs, which depend on compute time like everyone else. Prices matter.
A warm can of Diet Coke sat by the scanner.
Startups inside Fresno State’s Water, Energy and Technology Center also rely on compute to build prototypes and pilots. If hardware stays tight or cloud pricing bumps up with new rules, those costs land in grant budgets and timelines first. But not by much.
The policy piece to watch
Export policy has seesawed since 2023. Washington first restricted sales of Nvidia’s top-end accelerators, then suppliers introduced China-only variants, and later filings and briefings noted bans and licensing on those too. Reporting over the past year also described case-by-case approvals and revenue-share terms for some chip sales into China, a reminder that rules can shift again. Any of those moves can change what Valley institutions pay for GPUs or cloud time the next quarter.
The Shanghai meeting closed its opening session with delegates taking notes as Xi repeated his complaint about "overstretching" national security.
Central Valley AI is produced by the CVAI Newsdesk team and developed by Kaweah Tech, a regional firm that builds, deploys, and integrates AI solutions for businesses across California's Central Valley.