🌐 The Old Web Project · Bring back the human web · own your space, own your media, keep the web weird 🌐

A history of the web

From the personal web to the AI-mediated web

Three decades in one scroll — how the open, hand-built web became centralized and rented, and how the next one can be human again. Scroll to travel through time.

1990s · The Personal Web
1991

The web goes public

Tim Berners-Lee puts the first website online and gives the World Wide Web away for free — no license, no owner, no gatekeeper.

1994

GeoCities & the homepage era

Free hosting lets anyone build a personal “homepage.” Pages are organized into themed neighborhoods — the web as a place people live, not a feed they scroll.

1995

Webrings connect the frontier

Independent sites link together with prev / next / random navigation. Discovery is human and curiosity-driven — you travel the web instead of being fed it.

1996

Hand-coded, weird, and personal

Neon text, tiled backgrounds, guestbooks, hit counters, “under construction” GIFs, and 88×31 pixel buttons. Loud, imperfect, and unmistakably made by a person.

Early 2000s · Blogs, Forums & RSS
1999–2003

Blogging rises

Blogger, Movable Type, and WordPress make publishing easy while you still own the site. Anyone can have a voice on their own domain.

Early 2000s

Forums thrive

phpBB and vBulletin communities become deep, searchable archives of human knowledge — the kind that later vanish when they move to closed chat apps.

2005

RSS: following without a feed

RSS lets you subscribe directly to sites and read them in your own reader — no algorithm deciding what reaches you. The original, open “follow.”

Late 2000s · The Platforms Arrive
2003–2006

MySpace, Facebook, YouTube, Twitter

Social platforms make publishing effortless — and in exchange, identity shifts from websites you own to profiles you rent.

2009

GeoCities is deleted

Yahoo shuts down GeoCities, erasing millions of personal pages. Volunteers scramble to archive what they can. The lesson: when culture lives on a platform, it dies with it.

2010s · The Algorithmic Feed
2012

Feeds replace timelines

Instagram and others move from chronological to algorithmic feeds. A black box now decides what you see — and what your work reaches.

Mid 2010s

Organic reach collapses

Facebook Page reach is throttled toward zero. The audience you built becomes something you have to pay to reach. Creators start optimizing for the algorithm instead of for people.

2013

Google Reader shuts down

The most popular RSS reader is killed, nudging millions away from open subscriptions and toward closed, algorithmic feeds.

2020s · The Rented Web
2020s

Subscriptions & streaming everywhere

Movies, music, games, and software become rented access. Titles disappear from libraries; editions change silently; nothing is truly yours.

2023

The Reddit API shutdown

Sudden API pricing kills third-party apps like Apollo and a whole ecosystem of tools — a reminder that everything built on a platform is built on rented land.

2023–2025

Closed communities & link rot

Searchable forums give way to Discord servers no search engine can see. Domains expire, fan sites vanish, and the web quietly forgets.

2026+ · The AI-Mediated Web
2026+

Assistants answer; sites go unvisited

AI assistants summarize the web and hand back answers directly. Fewer people click through to the sites the knowledge came from — and centralization deepens.

The risk

Not just disappearing — unvisited

The danger isn’t only that websites disappear. It’s that people stop visiting websites at all — and the incentive to make them quietly dies.

What comes next · The Human Web
Now

The next web can be human again

Personal websites. RSS and directories. Self-hosting and open protocols. Physical media. Community-run spaces. Webrings. The tools to rebuild a web that feels like ours already exist — they just need people to use them.

You’re part of the next chapter

The story isn’t finished. Every independent page makes the web a little more human.