Ben Cera argues that near-frontier open models from China could push a 100M AI bill toward 5M and pressure frontier lab pricing. The post is a sharp market claim, not evidence of a crash, so the click is in the pricing argument and the missing measurements.