Skip to content
Deepak Gupta markTech Graveyard

Tech Graveyard/workflow

Google Reader and RSS (1999-Dying)

RSS gave readers control of the feed. Google Reader killed alternatives, then Google killed Google Reader. The whole category lost. AI agents may quietly bring it back.

Born 1999 · Died 2013 · Status: dying

Certificate of Death

Name of decedent

Google Reader and RSS

Born
1999
Died
2013
Age
14

Cause of death

Google Reader shutdown destroyed the category's center of gravity; social feeds replaced reader-controlled distribution

Survived by

Feedly, Inoreader, NetNewsWire, niche RSS readers, and AI news agents

Invented by

Netscape (RSS 0.90, 1999), then Dave Winer (RSS 2.0)

Status: DyingFinal breath: 2013

Filed by D. Gupta · guptadeepak.com

The hook

Google shut down Google Reader on July 1, 2013. RSS adoption never recovered. The dream of reader-controlled feeds lost to algorithmic social timelines, and a generation later, to AI-curated answers.

Thesis. RSS was the most successful failed standard in internet history. The protocol works. The technology is fine. The category lost because the economic model favored centralized social feeds. The deeper backend lesson: open protocols lose to closed networks even when the open protocol is better.

The story

The origin

1999. Netscape shipped RSS 0.90 as part of its My Netscape Network portal play. Dave Winer iterated to RSS 2.0 in 2002. The Atom standard followed in 2005 to clean up the format wars.

The early enthusiasm

2003 to 2010. Bloggers, journalists, and technical communities adopted RSS aggressively. Google Reader, launched October 2005, became the dominant interface. Bloglines, NetNewsWire, and Newsgator competed for share.

The Twitter shift

2009 to 2013. Twitter, Facebook, and Tumblr became the new follow-people interfaces. Algorithmic timelines replaced chronological feeds. Publishers chased the social audience and let RSS feeds languish or break.

The Reader shutdown

March 2013 Google announced. July 1, 2013 it shut down. The category's center of gravity disappeared. Feedly absorbed 3M users in two weeks but never recaptured Reader's scale. Alternatives exist; mass adoption does not.

The agent revival

2024 onwards. AI news agents (Arc Search, Perplexity Discover, custom Claude and ChatGPT workflows) implicitly do what RSS did. Reader-controlled feeds may quietly come back through AI, with the agent doing the polling and summarization that humans stopped doing.

Key data points

  • RSS 0.90: Netscape, 1999
  • RSS 2.0: Dave Winer, 2002
  • Google Reader launch: October 2005
  • Google Reader shutdown: July 1, 2013
  • Feedly user surge post-shutdown: 3M users in two weeks

Contrarian angle

The most successful person on Twitter and the most successful AI news agent are doing the same thing RSS did in 2005: aggregating sources you chose into a single feed. The pattern won. The protocol lost. The next generation of AI reader tools may rebuild RSS's vision under a different name, with the user once again in charge of the source list.

The flip side

What replaces it

The paired prediction in Future Tech.

Read the prediction

FAQ

Why did Google really kill Google Reader?

Officially: declining usage and Google+ priority. Practically: Google could not insert ads into a product where users controlled the feed, and Google+ required engineering attention. The strategic value of Reader to Google was negative.

Is RSS still useful today for any practical purpose?

Yes. Podcast distribution still runs on RSS. Many news sites still publish feeds. Tools like Feedly and Inoreader serve a smaller but committed audience.

Are AI news agents the new RSS?

Functionally similar, architecturally different. The agent acts as the user's aggregator and summarizer. The protocol underneath is HTTP scraping or vendor APIs, not RSS. The user-control pattern is back; the open standard is not.

What is the difference between RSS and Atom?

Both are XML feed formats. Atom (2005) is cleaner, better specified, and supports more metadata. RSS 2.0 is more widely deployed. Most readers handle both transparently.

Can decentralized social protocols succeed where RSS failed?

Mastodon and ATProto have better economic models (operator-funded servers, identity portability) than RSS did. Whether that translates to mass adoption is still being decided.

More from guptadeepak.com

Want the technical deep-dive on what replaces this?

Read the companion article

Related obituaries

More from the workflow graveyard.