The promises and challenges of web3

Here’s a primer on web3. A bit unstructured but hopefully it can help if you have a basic understanding of blockchain and crypto but are somewhat confused about the web3 term. Views are my own.

PS: I don’t have any strong feelings about the web3 term other than I think it’s an effective rebrand of crypto that’s helping many outsiders open their eyes to the tech and understand the concepts. Right now web3 is a convenient label for promising tech that has many flaws and is far from reaching its final form.

I see web3 as an evolution of the internet’s front end and a revolution and full rearchitecting of the back end.

What are the promises of web3?

Ownership

• Users, who were the product in web2, are elevated to stakeholders in web3 (although, stay vigilant or you might become the yield).
• Holding tokens of a protocol or an app lets users participate in its governance and own a piece of it.
• Tokens align network participants to work together toward network and token price growth.
• NFTs enable digitally-verifiable ownership of anything.
• Tokens fix web2’s core problem of all value being accumulated by a few companies.
• Token incentives are a powerful new web3 marketing technique that can lower customer acquisition costs. Tokens are self-marketing as holders become evangelists.
• Users become harder to track and target as they move away from web2’s social login schemes to anonymous or pseudonymous wallet-based app access, and as the web moves toward more transparent, cookieless data practices.
• Wallets become a ubiquitous user client in web3 and let users give selective read-only access to their data.
• Providing PII in web3 primarily happens on KYC-enabled front ends (of which there will be a lot though).
• There are however protocols aiming to tackle this new paradigm and help brands connect with web3 users.

Data portability

• Web3 makes value transfer a native function of the internet and allows it to move at the same speed as information in web2.
• Platform lock-in becomes a thing of the past as switching costs are low. Users can easily bring their data or content to a competing platform. At the same time, developers can deploy their applications across different networks.
• Users can have a persistent inventory of objects and carry it between apps. If their stuff increases in value, the user gets the upside. Objects can be traded in-app or on secondary markets.
• Creators take back IP ownership from the big platforms as they can sell their work directly to their audiences.
• Many front ends will be able to display your data but different from web2 they only provide a view into that data, they don’t hold it.
• For example, as a user, you have complete control over your data and the tokens/NFTs you own on a blockchain. Centralized services such as NFT platforms will let you access your data. The difference from web2 is the network effect doesn’t accrue to these services.
• Due to these features, platform fees/take rates are much lower than in web2.
• For the same reasons, web2 intermediaries with rent-seeking business models will be fewer. The higher the fees, the higher the risk of disruption for the incumbent.

Composability (and modularity)

• Open source and forkable software unlocks the power of composability.
• Composability lets developers use and reuse software components freely, mixing and matching them like Lego bricks.
• This lets developers reuse existing, reliable code freely in their own projects instead of having to reinvent the wheel.
• In crypto, we’re seeing composability in many areas, including:
• Public smart contracts that can be called by others.
• Apps that are cloned and integrated with each other.
• Protocols building on top of each other.
• Interoperable tokens.
• L1 atomic composability that lets you bundle several operations into a single transaction, essential to DeFi use cases.
• Some networks apply a modular architecture, separating the consensus and execution layers to solve scaling and security, and offer more flexibility for projects building on top of them.

What are the challenges of web3?

Onboarding and UX

• The concept of self-custody and understanding the consequences and responsibility of being your own bank is hard for new users.
• The learning curve is steep across the board, as you quickly have to deal with multiple networks and their limitations, tools and wallets, different address formats, and many unfamiliar terms and concepts.
• UX is still developer-driven and it shows.
• Too many front ends requiring KYC and a wallet can become a barrier to adoption.
• Some services bridge web2 with web3 with auth infrastructure to bring familiar social login processes to apps and wallets (not sure how I feel about this, they simplify UX but it’s still web2).
• Skeuomorphism is probably holding back real innovation in web3 apps and affecting how outsiders look at the technology (e.g., most people think NFTs are limited to art and collectibles, similar to how people once thought the web was limited to brochures and magazines).
• We also need more robust developer tools to facilitate the onboarding of more non-blockchain devs.

Interoperability

• Seamless cross-chain interop is needed to have total data portability and freedom.
• Users today are restrained to portability within individual networks or specific execution environments most of the time.
• This limits users to the individual networks’ capabilities.
• Both centralized and “trustless” bridges carry risks, attack vectors, and have poor UX.
• These limitations, along with high transaction fees, are a blocker for composability in practice.
• Blockchain data is public but usually painful to retrieve and manipulate.
• Developers building on multiple networks have to integrate and set up separate infrastructure for each of them. This gets even more complex when you need to have communication between the different networks.
• Deploying an app that connects with multiple networks through a single front end still requires a siloed approach and also complicates the UX.
• There isn’t yet an infrastructure stack that allows secure and easy cross-chain development.
• Devs transitioning from web2 are having a hard time, even more so when having to learn new, blockchain-specific languages.

Centralization

• L1s are a big part of web3’s infrastructure.
• Most L1s aren’t very decentralized.
• Not able to solve the trilemma, they sacrifice decentralization and security to scale.
• Most people don’t care yet.
• Building on top of centralized infra is not good.
• Solutions built on top of decentralized infra, however, and that rely on some centralized components are okay because the network effects still accrue as public goods and aren’t collected by a single company like in web2.

We should cover many more topics (e.g., identity, reputation, security, privacy, censorship, use cases, regulation), but that’s for another day.

I believe open-source, decentralized, and composable services one day can surpass the capabilities of closed, proprietary ones without compromising on UX and scalability. When the day comes we might see a decentralized network becoming the new App Store, followed by a phase where every company (thinks it) needs a dapp.

PS: No, web2 companies that start accepting crypto payments don’t magically become web3 companies. It’s positive that they integrate with existing solutions and help adoption that way but they aren’t building anything new.

PPS: a16z Crypto’s Chris Dixon has good thoughts on web3, and I recommend his posts for in-depth reads on many of these topics.