Silicon Valley's AI Civil War: Open vs. Closed Models and the Race to Regulate Superintelligence

Silicon Valley's biggest AI labs are publishing competing manifestos in Washington this week, and the core thesis dividing them is this: Does AI safety come from spreading powerful models to everyone, or locking them down? The fight is no longer theoretical — it's shaping legislation, infrastructure deals, and the global AI race.

Evidence For

• Kimi K3, a Chinese model with open weights anyone can download and modify, reached frontier-level performance — proving that...