A brief history of the internet, and why every web dev should know it.

Section · 01

Before the internet was the internet

In the 1960s the U.S. military used microwave transmission towers to relay communications. They worked well enough on a clear day, but they had two problems. Weather killed them. And if you took out one tower in the middle of a chain, the whole line went dark. After Sputnik launched in 1957 and the Cold War started looking like a real war, the Department of Defense decided it needed a network that kept working when parts of it got blown up.

Around the same time, university researchers were trying to share data between campuses and running into the same problem from the opposite direction — their connections kept dropping in the middle of long file transfers. The two groups met at a conference and decided to collaborate.

Section · 02

Arpanet — the first network

The Defense Department’s research wing (ARPA) built a decentralized network they called Arpanet. The key idea was packet switching: instead of one big continuous connection between two computers, break the data into small numbered chunks (packets), send each one through whatever route was available, and reassemble them on the other side. If one route died, the packets just took a different one.

The first message sent over Arpanet, in 1969, was the word login. The system crashed after the lo. That’s the entire history of software in one sentence.

Section · 03

TCP/IP — the standard everything runs on

Arpanet was a single network. By the 1970s lots of networks existed, and they couldn’t talk to each other. Vint Cerf and Bob Kahn proposed a common protocol any network could use: TCP/IP (Transmission Control Protocol / Internet Protocol). The idea was that each network kept doing its own thing internally, but spoke a shared language at the boundaries.

On January 1, 1983 — “flag day” — Arpanet switched to TCP/IP overnight. That’s the day the modern internet was born. Everything you do online today still moves over TCP/IP. The fact that your laptop, a Linux server in Frankfurt, and an iPhone all just work on the same network is because of this one protocol.

Section · 04

Web 1.0 — read-only

The internet existed for 20 years before most people had heard of it. What changed that was Tim Berners-Lee inventing the World Wide Web at CERN in 1989. He wrote three things: HTML (the markup), HTTP (the protocol for moving HTML around), and a browser to view it.

The first website, info.cern.ch, went live in 1991. Through most of the 1990s the web was Web 1.0: static pages, hand-coded HTML, no user accounts, no comments, no dynamic content. You read what someone published. You couldn’t write back.

Email — older than the web

Email predates the web by 20 years. Ray Tomlinson sent the first networked email in 1971 and picked the @ symbol to separate the user from the host. That little decision survived every protocol change since.

Section · 05

Web 2.0 — read-write

By the mid-2000s the web became interactive. Users stopped being readers and started being contributors. The shift had three triggers:

1. Cheap servers + faster connections (you could run a forum without a sysadmin team)
2. JavaScript got good enough to update part of a page without reloading
3. Companies figured out that user-generated content scales for free

Web 2.0 is Facebook, YouTube, Wikipedia, Reddit, Twitter — sites where the users arethe content. The technology underneath (HTTP, TCP/IP, HTML) didn’t change much. What changed was the culture and the business model.

Section · 06

Web 3.0 — semantic, decentralized, and AI

“Web 3.0” is a fuzzy term with three competing definitions, and you’ll hear all of them:

Semantic Web   — machines understand content, not just display it
Decentralized  — blockchain, crypto, ownership without platforms
AI-native      — agents that browse and act on the web for you

What it actually means in 2026 is some mix of all three. The practical version: APIs are first-class citizens, content is structured for machines (schema.org, JSON-LD), and AI assistants increasingly sit between you and the raw web.

Section · 07

Why the history matters

You can build a website without knowing any of this, but you can’t debug one. When DNS fails, when CORS breaks, when a port is blocked, when an SSL cert expires — every weird thing you’ll hit has a reason rooted in some 1970s design decision made by a guy with a beard at a university. The next lesson digs into how a single page load actually moves through that decades-old plumbing.