Receipts: A brief list of prominent articles proclaiming the death of the web (zeldman.com)

This piece compiles a short history of recurring proclamations that “the web is dead,” citing Wired cover pieces from 1997 (PointCast/push headlines), 2004 (RSS/Atom and XML syndication), and 2018 (apps eclipsing the browser). Each panic moment celebrated a new paradigm—push clients, reliable syndication standards, or standalone apps—as the death knell for the web, but the author argues these technologies instead augmented and reshaped the web rather than replacing it. For the AI/ML community the takeaway is practical: many recent claims that AI will “kill” the web repeat an old pattern of hype. Technically, standards like RSS, Atom and XML show how interoperability and syndication folded new capabilities into the web instead of supplanting it. More importantly, modern LLMs and retrieval systems ingest and depend on the web’s text, links and APIs; a scenario where AI destroys its primary data substrate is self‑defeating. The implication for researchers and builders is clear—investing in robust web standards, better content provenance, and retrieval-first architectures is a more realistic and productive response than predicting obsolescence. The web has historically absorbed and amplified new tech; expect the same with AI.