What is Web3?

In brief

Web3 is an umbrella term for a proposed next phase of the internet in which users interact with services through cryptographic accounts and decentralized protocols instead of relying entirely on centralized platforms to hold their identity, data, and assets. It is not a single protocol, a formal standard, or a piece of software anyone can point to; it is closer to a loosely shared set of goals and design preferences that different people apply differently, which is why disagreements about "what Web3 really is" are common and, to a real extent, unresolvable by looking up a specification.

Where the term came from

The specific "Web 3.0" framing used in crypto traces to a blog post by Ethereum co-founder Gavin Wood, published on April 17, 2014, titled "ĐApps: What Web 3.0 Looks Like." Wood described a "post-Snowden" vision of the web built on decentralized protocols and applications ("ĐApps") rather than centrally hosted services, motivated by concerns about surveillance and trust in intermediaries. This is a distinct usage from the earlier "Web 2.0," a term popularized by Tim O'Reilly's 2005 essay describing the shift toward user-generated content, social platforms, and web-based software, itself a description of a business and design shift rather than a blockchain concept. The "Web3" spelling and broader popular usage largely dates from later, particularly a wave of attention in 2021 as venture capital and mainstream commentary picked up the term; it is not the same project as Tim Berners-Lee's much earlier Semantic Web. Berners-Lee, the inventor of the World Wide Web and director of the W3C, used the label "Web 3.0" himself years before Wood's post, including in a 2009 W3C-hosted talk titled "Web 3.0 and Linked Data," to describe a web of machine-readable, interlinked data built on standards like RDF, a project centered on structured data and metadata rather than cryptographic ownership or blockchains. Readers should not treat these as the same idea just because they share a name.

The broad modern meaning

As currently used by its proponents, Web3 generally refers to an internet built around a cluster of related ideas: user-controlled cryptographic accounts instead of centrally issued logins, ownership of digital assets recorded on public blockchains instead of inside a single company's database, and applications built on decentralized protocols rather than solely on centrally operated servers. Ethereum.org's own explainer frames this as a shift from being a "consumer" of the internet with borrowed accounts on centralized platforms, to being an "owner" who holds assets and identity directly, without needing permission from an intermediary to participate.

It is worth being explicit that this is a project-affiliated description of an aspiration, not an independently verified account of how most deployed applications actually work today.

Ownership and control via cryptographic accounts

The mechanism proponents point to for this shift is the crypto wallet and the private key behind it: rather than a username and password controlled and revocable by a company, a Web3 account is, in principle, controlled only by whoever holds the corresponding private key, and any assets or permissions associated with that account move with it across different applications that choose to support the same underlying blockchain. Tokens function as one building block of this model, representing ownership, membership, or access rights that a user's account can hold independently of any single platform's cooperation. (See the companion articles on crypto wallets and cryptocurrency for how these mechanisms actually work.)

Decentralized protocols and relationship to Web1/Web2

Proponents typically describe Web3 as a corrective to a specific pattern seen in Web2: a small number of large platforms increasingly controlling identity, data, and monetization for content and services built on top of the open, comparatively decentralized Web1 network of independent sites. Web3's proposed answer is to move core functions, financial transactions, identity, and data storage or coordination, onto decentralized, permissionless protocols that no single company operates or can unilaterally shut down, with applications built as interchangeable layers on top rather than closed platforms.

Major criticisms and limitations

This vision has drawn substantial and specific criticism, much of it from technologists rather than from outside commentators dismissing the topic in general terms. Cryptographer and messaging-protocol developer Moxie Marlinspike wrote in a widely discussed January 2022 essay that, after directly testing a range of Web3 applications, he found that the practical experience still depended heavily on a small number of centralized infrastructure providers (companies running blockchain nodes and hosting front-end code on ordinary centralized servers), because most users and even most developers do not want to run their own infrastructure. His core argument was that protocols move much more slowly than platforms, which creates pressure for convenient, centralized points of control to persist even in nominally decentralized systems.

Public skepticism has also come from prominent technology figures outside the crypto industry specifically. In December 2021, Tesla's Elon Musk and Twitter co-founder Jack Dorsey publicly mocked Web3 as more of a "marketing buzzword" than a working reality, with Dorsey specifically arguing that Web3 would ultimately be owned and controlled by venture capital firms and their investors rather than by users, making it, in his view, a centralized system carrying a different label. These criticisms target different failure modes: Marlinspike's is a technical argument, backed by direct testing, about deployed infrastructure, while Musk and Dorsey's is a public, less technically specified argument about the concentration of investment and incentives around the term. Neither has been definitively resolved by developments since.

What proponents say in response

Wood himself has continued to describe Web3 as more of a sociopolitical project than a purely technical one. In a 2022 interview, he described his vision as an internet where services are provided in a peer-to-peer, algorithmic way rather than by a single company with "absolute power" over how a service works, framing it as a move away from having to trust the operators of centralized platforms. He also noted in that interview that his version of Web3 may not need to rely heavily on cryptocurrencies specifically, which is a narrower claim than much of the popular, token-heavy usage of the term.

Why there is no single agreed technical specification

Unlike a blockchain protocol, an improvement proposal, or a cryptographic standard, "Web3" has never been formalized into a single specification that a system can be checked against for compliance. It functions more like a label loosely covering decentralized identity, token-based ownership, decentralized applications, and decentralized finance, applied inconsistently across projects with very different architectures, some of which rely on decentralized infrastructure much more thoroughly than others. This is also why criticisms and defenses of "Web3" often talk past each other: a critique of one specific deployed application's centralized dependencies is not automatically a critique of the broader goals Wood or ethereum.org describe, and a description of the ideal is not evidence about how any particular product actually operates in practice. Readers should treat any specific "Web3" claim about a specific product as something to verify against that product's actual architecture, not as something guaranteed by the label.

What to learn next

The companion articles on blockchain, decentralization, cryptocurrency, and crypto wallets cover the individual technical building blocks this article's broader umbrella term rests on, and are a better starting point for evaluating any specific claim made about a "Web3" product.