What's this about?
An argument about what the web becomes once generated content is the majority of it — landing the same day as the 215,128 manufactured pages report and the Perplexity citation audit.
What is the specific mechanism to worry about?
Not volume but the collapse of the discovery layer. The web worked because search could rank human judgement, imperfectly, using links as a proxy for attention. Generated content breaks that proxy at scale — you can manufacture the signals — and the answer layer that replaced search inherits the problem, because it retrieves from the same poisoned index. The failure is not that bad content exists; it is that the mechanism for finding good content stops distinguishing.
Why is it worth noticing?
HN 71 points, 79 comments
- More comments than points, meaning disagreement rather than agreement.
- Read alongside the two Perplexity reports today, it stops being speculative.
- The retreat-to-books response in the manufactured-sources thread is people already acting on this.