China’s AI money question shifts from models to data centers and collateral

sankin_eth · x · 2026-07-24

• The post argues that after property and new energy, China may need a new sector to absorb newly created capital, and asks how AI can become financeable like real estate.
• It suggests large-model companies are hard to price directly, while data centers and cloud assets look more like traditional collateral-backed businesses.
• The author also says hardware manufacturing and future embodied AI fit China’s familiar industrial logic better than pure model companies.
• Despite that, the market still talks more about model companies than cloud or data-center plays, and the post speculates Chinese foundation-model firms may eventually need to list in Hong Kong.

Original post →

More from Infra

• Oracle cuts 21,000 jobs as Wisconsin OpenAI campus faces a $7B power guarantee hurdle — kimmonismus · 2026-07-24
• Leak claims MI500 adds optical interconnects and tops Rubin Ultra in key areas — zephyr_z9 · 2026-07-24
• Nunchaku Lite cuts peak VRAM by 50% and speeds image generation 30% in Diffusers — RisingSayak · 2026-07-24
• Open-source AI’s biggest week puts pressure on frontier labs and prices — a16z Podcast · 2026-07-24
• Hetzner Is Reportedly Moving Into LLM Inference — jonas_scholz · 2026-07-24
• AI governance chatter is cooling as automation and data plumbing keep rising — YvesMulkers · 2026-07-24

Infra channel →