The Vendor Bias Problem We Left Unresolved

Back in Story 6, I'd discovered our research was really vendor research. Citation validation caught the hallucinations beautifully but missed entirely that most valid sources were vendors selling solutions to the problems they were describing. I logged the issue and moved on to other things. Six stories of productive procrastination later, I finally built the fix: three-track architecture with half standard research, a quarter independent voices, a quarter contrarians. Vendor concentration went from majority to roughly a third, which turns out to be about as good as you can do without solving capitalism itself.