Beijing-based Moonshot AI released Kimi K3, a 2.8-trillion-parameter system that matches leading U.S. models on multiple benchmarks at lower cost. Independent verification of performance remains ongoing, and the model does not lead every category. Moonshot also plans to release a model for public download.
No left-leaning sources covered this
No center sources covered this
“China’s new AI just blew up Silicon Valley’s $1 trillion delusion”
Read at Washington Examiner →Kimi K3 undercuts Silicon Valley’s narrative that only lavishly funded U.S. firms can lead, showing competitive performance at lower cost and aligning with priorities of wider access.
“Challenges trillion-dollar valuations and rent-seeking logic in favor of deployment speed and affordability”
The release exposes fragile assumptions of permanent American dominance and turns AI into a national-security contest with an authoritarian rival.
“Emphasizes strategic competition, state backing, and risks of underestimating China”
Technological competition across borders drives progress and lowers costs, expanding individual access and reducing dependence on centralized labs.
“Prioritizes decentralized knowledge and user agency over corporate or state gatekeeping”
All three views accept the same selective framing of parameter count and price while ignoring incomplete verification, export-control issues, and training-data biases.
“Highlights groupthink around hype metrics and failure to interrogate source qualifiers or geopolitical realities”
Ratings by MBFC