The Internet
A “network of networks” made up of fiber-optic cables, routers, computers, and protocols such as TCP/IP. It allows different devices to exchange data.
From zero to Ethereum · 01 / 01
Web1, Web2, and Web3 are not three versions that replaced one another overnight. They are overlapping ways of organizing the web: who publishes information, who records the facts, and who sets the rules.
The difference is not how the page looks. It is who has the final say over the records that matter.
Once you have this frame of reference, blockchain, wallets, smart contracts, and Ethereum become more than slogans.Before discussing Web1, Web2, and Web3, separate two terms that are often confused: the Internet is the infrastructure; the Web is the system of information and applications that runs on top of it.
A “network of networks” made up of fiber-optic cables, routers, computers, and protocols such as TCP/IP. It allows different devices to exchange data.
The space of information and applications that runs on the Internet, usually accessed through URLs, HTTP, HTML, and web browsers.
Email also uses the Internet, but email is not the Web. Likewise, Web1, Web2, and Web3 are not three versions of TCP/IP, and none has an official worldwide release date. They are conceptual labels that help us examine how online applications have evolved. ethereum.org also presents these periods as broad teaching categories and notes that Web3 has no single, rigid definition.[3]
Are ordinary users limited to reading, or can they create, interact, and deploy their own programs?
Do the facts live on one server, in one company's database, or in state verified by many participants?
Who can ban an account, change a balance, revoke access, or upgrade the rules the system executes?
The deepest difference among the three Web eras is not the three verbs “read, write, own.” It is this: who records the facts in the system, and what guarantees users’ rights?
The most common mnemonic is: Web1: read; Web2: read and write; Web3: read, write, and own. It is useful—but only with careful qualifications, so that history does not collapse into a slogan.
Sites publish content and visitors primarily read it; open protocols and hyperlinks let information flow around the world.
Users create, socialize, and transact; large platforms package accounts, recommendations, payments, and data into accessible services.
Wallets, blockchains, and smart contracts seek to prevent key state from being determined by a single platform alone.
The arrows represent capabilities accumulating over time, not one generation deleting the last. Many Web1-style personal sites still exist today. Web2 platforms remain a dominant part of the Internet. And most Web3 applications still use browsers, cloud services, and the interaction patterns and user experience associated with Web2.
| Dimension | Web1 | Web2 | Web3 |
|---|---|---|---|
| Primary activities | Reading, browsing, and following links. | Creating, interacting, collaborating, and transacting. | Signing transactions, verifying state, holding assets, and making programmable exchanges. |
| Identity and access | Visitors or simple accounts. | Platform accounts and social logins. | Wallet addresses and verifiable credentials. |
| Authoritative record | Pages and files stored on the site's server. | A platform database controlled by a company. | Onchain state collectively verified under public rules. |
| Payment rails | Usually relies on channels outside the Web. | Integrates banks, card networks, and payment platforms. | Protocol-native assets and settlement. |
| Main advantage | Open standards, independent sites, and free linking. | Usability, efficiency, real-time interaction, and scale. | Open participation, verifiability, and composability. |
| Main trade-off | Higher barriers to publishing and interaction. | Accounts, data, and rules are centralized. | Key-management responsibility, fees, performance constraints, and governance complexity. |
The dominant pattern of the early Web was simple: site operators published pages, visitors read them in a browser, and hyperlinks led from one independent site to another.
Tim Berners-Lee proposed the World Wide Web at CERN in 1989, then helped build its early server, browser, and core specifications. URLs, HTTP, and HTML gradually became open standards, allowing independently operated sites to connect into a global information space.[1][2]
The “read-only Web” is a summary of the dominant experience, not a technical prohibition. Guestbooks, forms, forums, chat rooms, and e-commerce all existed early on. The real difference is that most ordinary users were not yet continuous contributors on large platforms, and interactive features had not been packaged into standardized services as they are today.
The defining property of digital information is that it can be copied without consuming or altering the original. You can send an article to one hundred people without losing your own copy. The Web drove the cost of copying, finding, and distributing information dramatically lower—a profound advance for open knowledge.
But that same property reveals another problem. If a digital object must remain scarce, or be spent by only one person, copying a file cannot by itself express a transfer of control. Information networks are good at answering “What is this content?” They do not inherently answer “Who controls it now?”
Web1 is like a global digital library: open protocols connect many bookshelves, but that model provides no internet-native system of property rights and settlement.
Web2 turned ordinary users from readers into active participants. Posting, commenting, liking, uploading videos, shopping, hailing rides, collaborating at work, and playing online games all became part of daily life.
Platforms solved a long list of difficult problems for users: account management, data storage, content moderation, search and recommendations, instant messaging, payment integration, spam protection, and password recovery. By delivering a stable, convenient, end-to-end experience, they created enormous network effects.
This makes the platform database the system’s ultimate source of truth. Your 8,420 followers on a social network or 100 gold coins in a game are, technically, usually records in a company’s database. Under its terms of service, the platform can change its interfaces, reorder results, restrict transactions, suspend accounts, or discontinue the service.
Low barriers to entry, high throughput, privacy controls, content moderation, customer support, and account recovery let billions of people participate safely and conveniently.
Platforms usually control identities, data, recommendations, payment interfaces, and access rules unilaterally. Users can rarely leave with their complete relationships and records intact.
Centralization is not inherently malicious. It is an organizational model that uses a clearly identified operator and a unified database to gain speed, accountability, and ease of use. What matters is recognizing the exchange: you receive a service, and you also give its provider final authority over the records.
Web2 resembles a well-run shopping mall: it is easy to set up shop and reach customers, but the operator owns the membership system, the premises, and the rules of business.
Web3’s central experiment is not to give websites a new look. It is to keep certain key records from existing solely in one company’s database.
In this course, Web3 means the Web in a blockchain context: users initiate and authorize actions, while nodes in a network apply public rules to determine whether records may be updated. Later lessons on cryptography, transactions, and blockchains will explain how authorization, verification, and consensus work in practice. The term remains contested, and it should not be confused with “Web 3.0” or the Semantic Web—the W3C usage associated with machine-readable data.[4]
State is the set of relevant records a system recognizes at a given moment—for example, balances, controllers, or membership status. A valid action does not send a physical object through a cable. It updates the records the system recognizes from “before the action” to “after the action.” For now, this intuition is enough; transaction structures and cryptographic details can wait.
Here, value means more than market price. It broadly includes digital rights and entitlements that can be explicitly controlled, transferred, verified, and programmed: native assets, tokens, memberships, game items, claims, lending positions, or voting rights.
The system can clearly distinguish the object, its quantity, and its current state.
Public rules can determine who is entitled to authorize the next action.
Copying data does not also copy the control recognized by the protocol.
A valid transaction can update the system’s state so control moves from one party to another.
Third parties can inspect key records without needing access to a platform’s internal systems.
Transfers and use can carry conditions that code executes automatically.
Web3 does not eliminate trust. It attempts to make key rules public, verifiable, and resistant to unilateral changes by a single operator.
Imagine earning a “golden sword” in an online game. It may look identical on screen, yet the real control relationship changes completely depending on the system behind it.
A webpage presents the image, attributes, and story; the server provides the information.
A company database assigns the item to your game account.
A smart contract records your address as the token’s controller, and transferring it requires authorization.
Even if the third model is fully implemented, the sword does not automatically work in every game. Another system must still understand its standard, artwork, statistics, and gameplay. A portable protocol record does not create product compatibility by itself.
The image itself also remains copyable. A blockchain can establish which address controls which token within protocol state; it does not make the image impossible to copy or automatically determine copyright ownership.
Real products rarely belong to only one generation of the Web. They often combine open webpages, platform services, and shared networks, so you must keep asking which layer holds each key record.
Browsers, links, and public information
Inherited from Web1Accounts, search, recommendations, notifications, and large files
Inherited from Web2Key state that multiple participants must verify together
What Web3 attemptsWeb3 applications still use the open protocols established by Web1 and draw heavily on Web2 interaction patterns, cloud services, and product design. They do not build a separate internet from scratch; they add new coordination layers—such as wallets, blockchains, and smart contracts—on top of the existing network.
Web1 expanded access to information. Web2 expanded participation. Web3 attempts to expand verifiable control and coordination.
Do not memorize three labels in isolation. Follow the question “Who records the facts?” and you can see why studying Web3 leads us to blockchains and Ethereum.
Open standards and hyperlinks let independent sites publish, copy, and connect content for a global audience.
Unified accounts, databases, and services made creation, collaboration, payment, and real-time interaction easy.
Key accounts, relationships, balances, and rules became concentrated in platform databases, where the platform retained final authority to change them.
Signatures, blockchains, and smart contracts allow some records to be jointly verified by many participants under public rules.
Can a publicly verifiable shared ledger be extended into a publicly verifiable state machine that executes general-purpose rules?
Web1 let information circulate openly. Web2 let platforms organize interaction efficiently. Web3 attempts to let actions over digital rights and key state be authorized under public rules, with the resulting state verified and made programmable on open networks.
Lesson complete