Opinion: Blockchain applications should be a niche subset of Web2, rather than replacing it as Web3
TechFlow Selected TechFlow Selected
Opinion: Blockchain applications should be a niche subset of Web2, rather than replacing it as Web3
Blockchain applications should be a niche subset of "Web2," not something new, nor something that can or should "replace" "Web2."
Navigating Web3 tides with focused insightsWritten by: Polynya
Translated by: TechFlow
In cryptocurrency, a common belief is "we must decentralize everything and put it on the blockchain," leading to odd memes like "Web3."
In reality, public blockchains are only suitable for very specific use cases—when both of these criteria are met:
- Peer-to-peer operations: Software that anyone can run, with information passed directly between participants;
- Strict global consensus: Everyone must agree on exactly the same outcome;
Thus, blockchain applications should be a niche subset of "Web2," not something entirely new, nor something capable of or meant to "replace" Web2.
First, not everything needs to be decentralized. Most services are best run in a centralized manner. Moreover, even in cases where decentralization is beneficial, peer-to-peer operation alone is often sufficient. In other cases, you only need local consensus—global consensus is not only unnecessary but wasteful.
A case where decentralization is highly beneficial yet achievable without a blockchain: the Alliance for Open Media. Background: A private company, MPEG-LA, demands licensing fees for all media content—music, video, etc.—on the internet. While open-source alternatives do exist, they’ve never gained adoption due to lack of infrastructure. Therefore, a coalition of companies and independent researchers collaborated to create an open standard for everyone—AV1 and AVIF: a remarkable public good. Global consensus isn’t needed here; local consensus within the alliance suffices, essentially forming a real-world decentralized organization.
So far, only two applications simultaneously require both criteria: money and identity. Now, it’s true that many monetary and identity applications only require local consensus, which traditional systems already provide—though they could be enhanced using technologies like smart contracts.
Of course, money can take various forms—non-sovereign currencies, representations of real-world assets, governance/ownership tokens, collectibles, loyalty rewards, social tokens, etc.—and so can identity. There are also interesting applications that combine money and identity.
However, in most cases, putting most things on a blockchain is wasteful. Hybrid solutions are possible—what I previously called "cedec" (centralized/decentralized) in an earlier article.
Let’s consider a "metaverse" scenario, or simply an open-source massively multiplayer game: a 3D virtual world where people participate. Notably, this isn’t a traditional game made by a company like Blizzard or Bungie. For traditionally developed games, only local consensus is required. Thus, a consortium of game studios, engines, and service providers can establish standards and operate them, though these elements cannot extend beyond the game itself.
Now imagine a new type of game, developed openly with many forks. Here, 3D rendering (locally or via cloud streaming), zero-knowledge proof generation, physics engine calculations, etc., can mostly or entirely happen peer-to-peer. You only need anti-cheat software to maintain local consensus—no blockchain required.
However, each user could have a global identity. Governance rights could be granted to players, revenue shared with developers, and certain in-game elements—perhaps time-limited cooperative puzzles involving physics or narrative—could require strict global consensus.
Let’s push the hybrid model further: a centralized company runs a fork of the above decentralized game but still wants to participate in the broader ecosystem.
So, the game primarily operates locally, in the cloud, or peer-to-peer, with only select components requiring blockchain-like solutions. The company runs its application instance, keeps data centralized, but chooses to publish high-value transactions and assets—such as collectibles under the ERC-721 standard—onto a committee layer, or even full rollup guarantees on an L1 data layer. Crucially, even with centralized data storage, the worst-case scenario becomes asset freezing—not seizure or theft.
This way, you achieve a hybrid solution that maximizes the strengths of each component and mini-use case in the most efficient way possible.
Outside the realms of money and identity, are there entirely new experiences that are both possible and meaningful?
Certainly. But so far, I haven’t seen anything beyond these. Using public blockchains for other purposes is wasteful. What I *have* seen are innovative combinations of the two—even when products are mainly centralized or achieve decentralization through non-blockchain methods—enabling genuinely new use cases.
Join TechFlow official community to stay tuned
Telegram:https://t.me/TechFlowDaily
X (Twitter):https://x.com/TechFlowPost
X (Twitter) EN:https://x.com/BlockFlow_News
