The trust collapse: Infinite AI content is awful (arnon.dk)

AI’s ability to generate content at near-zero cost has created an unintended collapse of trust: prospects—especially in B2B—now ignore outreach because they can’t tell real human engagement from mass-produced AI slop. The essay’s concrete examples make the point: a $10/month Claude license and near-free generation means anyone can produce polished, “personalized” emails, driving outreach volume from ~10/week to ~200/week while a prospect’s ability to detect authenticity drops (pattern recognition from ~80% to ~20%). The result is a broken signal-to-noise ratio where the cognitive cost of verifying messages exceeds the expected value of engagement, so recipients simply trash most inbound contact. For AI/ML companies and marketers this shifts the game from marketing funnels to “trust funnels.” Buyers increasingly ask not whether a product works, but whether the vendor will still exist, be different, and care long-term (unit economics, survivability, real human support). The remedy is human-led trust building: use AI to improve relevance and segmentation and to flag high-value leads, but invest real humans to create demonstrable credibility, transparency (metrics, business model, customer outcomes), and ongoing advocacy. Brands that treat trust as a strategic, long-term asset rather than automatable volume will stand out and reclaim real engagement.