>The frontier labs have already lost the war for coding

This is a delusional take. Sorry, but anyone claiming this hasn't used Fable and compared it to the current best open source models. I see a lot of hype posting about GLM5.2. I see absolutely ZERO people using it in production compared to GPT 5.5 or Opus 4.8.

Coding agents are edging into diminishing returns for common tasks. The whole Opus 4.5+ arc shows this for a large swathe of people. Chinese Fable is likely <=6 months away. US Frontier labs are structurally disadvantaged in the long term so advantage is only going to skew China.