The internet is not what you see when navigating online—that's the World Wide Web, which is supported by hardware called the internet.

Resembling a roadway with traffic laws that all cars must follow, the internet is the hardware that connects a source to a destination, and these connections must follow specific software rules. Although data had to travel through about 10 networks to reach a destination in the past, data centers increasingly consolidate traffic, just as a highway does for cars, to improve data speeds.

The World Wide Web began as read-only, static webpages on Web 1.0, where data could not be entered. With Web 2.0, the online interface became a read-write space for data entry, post creation, and forum participation, among other activities. Web3 is intended to be decentralized, user-owned, and built on blockchain technology.

Invented by Tim Berners-Lee in 1989 while working at CERN—home of the world's largest particle physics laboratory—the Web was developed to facilitate the sharing of information among scientists at institutions worldwide. By the end of 1990, the first Web server was operational as a proof-of-concept on a NeXT workstation.

The internet was officially born in 1983 and had more than 150 million global users by 1997, creating an international web of seemingly separate networks. This data visualization shows the growth of the internet from 1997 to 2021. Data in this...

The "surface web," consisting of billions of public webpages, excludes nonindexed digital content, such as corporate intranets and email inboxes. A deeper layer, called the dark web, is also hidden and requires special software to access.

DNS translates domain names, such as join1440.com, into numerical IP addresses, such as 192.0.2.1, which computer systems use to determine what content a user is looking for across the World Wide Web, making navigation easier for the vast majority of users. Web outages can occur if the DNS servers making these translations become inaccessible.

Through an open, transparent, and collaborative model built on open standards development and shared global ownership, the internet is kept freely accessible amid evolving international policies. Key organizations, such as IANA, which allocates IP addresses and fulfills administrative functions for the Domain Name System, function as non-profit organizations.

While TCP/IP sends data in packets to the correct destination and checks for errors, UDP omits such safety features to speed up data transfer for real-time communication and streaming. Software rules such as HTTP and HTTPS enable the transfer of content on the World Wide Web, while SMTP handles email messages between servers.

The system consists of several layers—application, transport, internet, and network—each responsible for specific tasks. These include interacting with a web browser, checking for errors, sending packets of data across various networks to avoid congestion on their way to a destination address, and managing the transmission of information across hardware, such as an Ethernet cable.

The surface web (5%-10%) consists of indexed sites—those traditional search engines have visited and stored so users can discover them via web search. The remaining 90%-95% makes up the deep web, which is not indexed and includes email and banking portals that require login credentials. The dark web is unindexed and requires anonymizing software to visit.

Web3 is the idea of a new, decentralized internet built on blockchain technology. It aims to shift control from centralized entities like big tech companies to users through distributed ledgers and smart contracts. Key components include cryptocurrencies, NFTs, and decentralized autonomous organizations (DAOs) that operate without a central authority. Web3 promises more user privacy, ownership, and control over data and digital assets.

Web1 (existing roughly from 1990 to 2005) was about decentralized and community-governed open protocols. Most of the value accrued to the edges of the network: users and builders. Web2 (roughly 2005 through 2020) was about siloed, centralized services run by corporations.

Many pages on the dark web are not publicly broadcast and are available only to relevant parties on a need-to-know basis. Although some online directories list the URLs for these webpages, connecting to them still requires specialized software, as these pages block connections from traditional web browsers such as Chrome.

Blockchain? Crypto? NFTs? Oh my. Gilles DC provides a simple breakdown of what Web3 is, what it involves, its connection with web currencies, and how you can begin your own explorations of this brave new world. Gilles promises to describe Web3 in one take, explain how the reader can take advantage of it, and how blockchain might build a better society as a result.

Within the Deep Web is the small and hidden Dark Web, which is inaccessible via traditional web browsers. Dark Web inhabitants are anonymous and use special software to access encrypted websites that can assist with both free speech and criminal activity.

Founded in 1996, the Archive promotes open access and the digital preservation of online information and, as of 2026, contains more than 200 petabytes of content, including over one trillion web pages, 56 million books and texts, 13 million audio recordings, 1.3 million software programs, and other digital content.

Subsea cables carry the majority of data sent across the globe on the internet. An estimated 552 such cables—with more in development—lay on the ocean floor, their thickness similar to a garden hose, made up of optic fibers covered in copper.

Mosaic was a web browser that allowed people to more easily access webpages on the World Wide Web. But unlike previous browsers, Mosaic was a graphical browser, which means it displayed images alongside text. This helped turn the web from a place for scientists into a cultural location that captured people's attention.

Crawlers are programs that automatically browse the internet for links and index pages for search engines to pull up when page content matches search queries. Networks of global data centers also create copies of content as fail-safes for potential failures to ensure access isn't ever lost.

Since the inception of the internet as a tool for connecting teams of researchers, technological and infrastructure advancements have progressively increased connectivity. Connections between computers, mobile devices, all devices, and now AI agents will be succeeded by an internet that integrates perception and the senses toward an eventual quantum internet.

While unlimited storage providers exist, this does not mean online storage is limitless, as it relies on servers, cables, power, and physical space, all of which are finite. However, the cloud seems infinite because the number of servers and data centers built outpaces demand. Storage solutions are also routinely replaced with more efficient options.

In 2005, 21-year-old Alex Tew set out to make $1M by selling $1 pixels on his website, effectively creating a standalone internet billboard. The gambit worked, and because the advertisements were one-time payments, the site looks the same as it did in 2005, despite many of the advertised sites going defunct since.

During a drive-by download attack, security flaws in a program or operating system (e.g., Windows) used to access online content are exploited, and malicious code is downloaded—for example, when loading a hacked webpage—onto a user's device without their knowledge. Restricting internet browsing to safe sites and installing security updates helps prevent these attacks.

NSFNET was launched in 1986 by the National Science Foundation as the first network available to all researchers, connecting them to a system of supercomputer facilities. Despite growing from 2,000 computers in 1985 to 2 million by 1993, the rapid expansion of commercial internet services led to NSFNET's decommissioning by 1995.

In the early days of the World Wide Web, users needed to type URLs directly in the search bar to navigate to webpages (as opposed to the more general keyword search that browsers use now). This made simple, short, and intuitive domain names a hot commodity. So having a domain name, like Pets.com or Rent.com, that was easy to find and understand was like owning prime real estate on the WWW.

These connections between everyday objects, each other, and data processing centers enable devices to adapt to user needs and improve their ability to respond to real-time data. IoT networks are designed to enhance the efficiency of homes, cities, and industries.

Browsing the dark web requires onion routing, which prevents visitors from being identified and the locations from which they connect from being traced. This has been useful to activists in environments affected by political oppression or instability. However, those unfamiliar with navigating the space may encounter malware and network attacks from denizen hackers.

Although encryption existed to mask the content of communications across the internet before the 1990s, those managing overseas network infrastructure could determine a spy's affiliation based on where they saw data sent to and received from. To keep this information secret, the US Naval Research Laboratory developed onion routing, which enabled anonymized web browsing. The release of the technology to the public obscured sensitive communications further by growing the volume of anonymous data.

Over 750,000 miles of fiber-optic cables run along the seafloor, serving as a freeway for the Internet's data. Despite subsea cable projects costing hundreds of millions of dollars, depending on the length of the cable, it remains the fastest and most cost-effective way to send information.

The IoT enables physical devices to share data collected via sensors or specialized software, often without human intervention. Controlling the air conditioning and heating in one's home via an app, as well as receiving alerts when away from home about someone at the door, is made possible by the IoT.

A mix of centralized cloud infrastructure and legacy protocols leaves the internet susceptible to downtime from minor cyberattacks or software bugs, infrastructure damage caused by extreme weather, and human error if a DNS service provider—the internet's phone book—or a backend system, such as Amazon Web Services, were compromised.

Various countries around the world, including China, Russia, and Iran, control their local DNS infrastructure, allowing them to block access to online content by preventing the internet "phone book" from finding the IP addresses for specific URLs. Government-controlled internet service providers can also cut off access entirely or throttle specific traffic.

Running between the US, UK, and Spain, the cable is part of an extensive underwater network of fiber optic connections that transmit data across the world at light speed. Equipped with 16 fiber pairs, the private cable was Google's first-ever subsea connection to Spain.

America Online's homepage was the most popular website through mid-2021, when the Yahoo homepage took over and, shortly afterward, became the first website to reach 1 billion monthly visits. It would not be dethroned until early 2006 by Google, a major player in the search engine market.

Since the latest version of the onion routing protocol, dark web domain names consist of a random set of 56 alphanumeric characters followed by .onion (e.g., propublica.org → p53lf57qovyuvwsc6xnrppyply3vtqm7l6pcobkmyqsiofyeznfu5uqd.onion). The last character is always "d" due to the encryption algorithm's requirements.

ISPs shouldn’t be able to block some sorts of data and prioritize others. Here’s what to know about the struggle to treat all information on the internet the same.

The dead internet theory proposes that bots create most online content. These bots are believed to create social media posts to boost engagement and increase advertising revenue. However, the accounts doing the liking and commenting are also believed to be bots. The theory suggests governments use bots to manipulate public opinion.

Although ARPANET connected various national laboratories and supercomputing facilities by the 1970s to facilitate scientific collaboration, the system was slow and required highly trained operators. The launch of NSFNET in 1986 provided a more flexible and powerful network that went on to connect over 170 academic and research communities and link to its European counterpart, EASInet.

The Whole Earth Catalog, a counterculture guide to alternative living, was a repository for hippie ideas. In 1985, its founder decided to bring Whole Earth to the nascent internet with the Whole Earth ‘Lectronic Link, or WELL, an early web forum.

The Submarine Cable Map provides users with data from TeleGeography's Transport Networks Research Service on individual submarine cables, including cable owners, suppliers, URLs, landing points, lengths, and notes of interest.

From wooden machines to wearable tech, computers have changed a lot in appearance, but they all still perform the same essential tasks. Early computers were made of wood and metal—some were so big that they filled entire rooms just to solve simple math problems.

On July 29, 1858, two steam-powered battleships met in the middle of the Atlantic Ocean. There, they connected two ends of a 4,000 kilometer (2,500 mile) long, 1.5 centimeter (0.6 inch) wide cable, linking for the first time the European and North American continents by telegraph.

Being able to track how users behave on websites is a massive opportunity for companies to hone their products, target their ads, and more. Dozens of trackers are on the average website you visit, depending on how you deal with that site's cookie policy, and Google is the chief tracker of them all. Learn about these companies and what they are tracking with this informative blog visualization.

Explore archived images of homepages on websites dating all the way back to 1991.

As long as users do not pay for illegal goods or services or acquire illegal content, they cannot be prosecuted. Those on the dark web include academics, researchers, and cybersecurity experts gathering threat intelligence on cybercriminals, cyberterrorists, and state actors, as well as protesters, activists, and grassroots organizers.

Encrypted networks allow users with internet access anywhere in the world to view blocked content and avoid surveillance when communicating evidence of abuse to independent journalists. The same networks can support democracies by helping activists coordinate protests on dark web forums.

And yet, the mysterious man who owns Milk.com likely isn't selling his valuable site any time soon.