Web3: fad or future?

Recently there has been a lot of controversy surrounding web3. It all started with the following tweet from Jack Dorsey. His claim being that web3 is largely owned by a16z. Soon enough, Elon Musk joined him. And you know when two of the biggest Twitter thought-leaders pick a fight. Everyone and their moms are going to chime in with their thoughts. And this is what inevitably happened. Tweetstorm below by Aaron Levie (CEO of Box) criticizing web3: Attack on venture capitalism by CEOs led to some arguments of defense. Fred Wilson who is the famous VC of Union Square Ventures wrote an article for thesis of web3 titled Why web3. In which, he invited criticism saying that debate is healthy, but he believes that there is sufficient evidence to believe that the movement will bore fruit — as some of it is already in action — particularly praising the novelty of open databases. Balaji tried to explain how the traditional funding mechanism is broken and how aspects of web3 solves it by essentially creating a transparent on-chain cap table that he coined mirrortable. In a way, he is saying that capitalistic aspects of web3 are inherent features rather than a cause or an effect. History and motivations One can argue that Web3 is a movement, not an architecture, nor a product. And in order to understand it, one should go back to the earliest signs of it. A git commit introducing a JavaScript library called web3.js. Gavin Wood is attributed for coining the term web3. Although, the term itself isn’t new (see Semantic Web). The earliest blogpost on the Ethereum forum describes web3 as an effort to claim back the web using decentralized technologies for storage, messaging and communications. The author writes: Technologies like Swarm could serve as the underlying static hosting infrastructure, removing the need to highly distribute and cache specific content. Because “decentralized dropbox” has been discussed with such frequency, expect HTTP-like bindings or services to be built atop this type of blob storage, making integration with the decentralized web 3.0 even simpler. This effort will also allow replacement of typical content delivery networks (CDN) with a distributed hash table (DHT) pointing to file blobs, much how BitTorrent works. Because of the flexibility offered by ethereum contracts, the model of content access could be creator pays, reader pays, or some hybrid system. So we’ve just replaced the need to have caches, reverse proxies, CDNs, load balancers, and the like to serve static content to users. Another way in which Etheruem could impact this traditional infrastructure is by replacing business logic application tiers with on-blockchain contracts. Chris Dixon, another big proponent of decentralized web, critiqued in his post why decentralization matters that in the current design of web applications, the only way to continue growing for a software company after an initial growth stage is by extracting data from users and competing with complements over audiences and profits. This leads to distaste and a desire to look for alternatives. Vitalik has a more technical spin on decentralization and why it matters. He listed three main arguments for decentralization in his post meaning of decentralization: Fault tolerance — decentralized systems are less likely to fail accidentally because they rely on many separate components Attack resistance — decentralized systems are more expensive to attack and destroy or manipulate because they lack sensitive central points that can be attacked at much lower cost than the economic size of the surrounding system. Collusion resistance — it is much harder for participants in decentralized systems to collude to act in ways that benefit them at the expense of other participants The architecture Enough on philosophical discussions about design. What tangible product do we really have? Gavid Wood’s original proposal of using Smart Contract, Swarm and Whisper protocols hasn’t really materialized as he would have liked; either due to poor technical design, lack of adoption or lacking incentives. And to my knowledge, none of the top DApps (decentralized apps) are fully decentralized (both technically and ownership-wise). And maybe, Jack’s criticism is not unwarranted as a recent study indicated that 10% of buyer–seller NFT pairs have the same volume as the remaining 90%. Does this mean the movement has failed? Not quite. But it has drifted a little from its original intentions and evolved to better to suit the current needs. This evolution is expected and should not come as a surprise. If anything, it is a sign that this is growing more powerful. Pretthi wrote a great piece on the architecture of a web3.0 application that can achieve its original ambitions of being maximally decentralized. In my opinion though, this level of decentralization is not necessarily needed and if pursued we will be blocked by some the constraints of new decentralized architecture as some components of the architecture are either lacking in scalability or security. I think a more realistic scenario is when Web2 architecture morphs into a Web3 architecture. Consider this scenario by Qiao Wang: I personally think that this scenario is highly unlikely, simply because a company as large as Facebook (or Meta) will unlikely be able to pivot to this extent. It does make a compelling argument thought. The existing web application landscape is on a trajectory to adopt crypto primitives. The bigger question is how will that manifest? Growth of the ecosystem From 500K monthly active users in June 2020 to 10 million in August 2021, MetaMask has grown over 1,800% in just one year. Compare that with the growth of Facebook in the early days. I don’t think the comparison is fair since one is social media network while the other is a tool. But the point remains: the user adoption figure is staggering! But it even more important that this isn’t a UX that everyone is already comfortable with. This is a learned behavior. You get the MetaMask extension from Chrome Extension Store (not your usual app store) and then you create a secret phrase and generate a cryptographic key pair which is used as means of authentication to all your DApps. The sheer fact that so many people are doing it means that we are exploring new horizons. And folks aren’t afraid to try new things (habit fueled by incentives rather than curiosity). The core tenant of any web3 project is a token. And there are two main formats: Fungible tokens (used for trading and attributing value) Non-fungible tokens (used for describing a solo asset) NFT marketplace OpenSea has grown at remarkable speed this year. So have alternative exchanges and yield generating applications. Aspects of Governance Remember the Cambridge Analytica scandal? The GDPR privacy rules? Mark Zuckerburg’s hearing? Donald Trump’s Twitter ban? They all have something in common. These are all questions of governance. … And the people need to have a say in how they want these platforms to be run. Fred Ehrsam outlines two critical components of governance in his post on programming our future: Incentives: Organisms are biased towards their own survival. This commonly manifests in changes to the reward structure, monetary policy, or balances of power. Mechanisms for coordination: If one group can coordinate better than another it creates power imbalances in their favor. He then goes on draw an analogy with how Bitcoin coordination system works. In today’s governance model, the key idea is to allow token holders to vote on board decisions and profit from dividends when issues. Examples for such decisions are: Trading fees Asset listings Risk parameters Protocol inflation schedules This is typically complemented with discussion forum where community engages in a dialog. Most protocol tokens have a defined reward structure that can be used to form real PE ratio (Price-to-Earnings) formulas. And while, some tokens still don’t have such reward structure, folks do expect them to come up with one soon. So, we have discussed various features of a web3 application, however, we said that the movement originally started with its mission to decentralize the web and give ownership back to the people. How do we fare? Centralization in the world of DApps Are we really working towards decentralization? Is this movement even going to work? Moxie (creator of Signal and a famous cryptographer) thinks that it is reminiscent of web1 more than web3. Moxie criticized the decentralization ambitions of web3. Stating that centralization on the web happened for a good reason: No one wants to run their own servers Decentralized technology (particularly protocols) changes slowly He drew an analogy to the email servers, the SMTP protocol and dominance of centralized services such as Gmail. He dabbled in the space by creating a DApp that allowed user to track an underlying NFT (a neat idea I must admit). Couple of lessons learnt in this process: Besides the state of the application nothing is particularly distributed Servers are still the critical part of the architecture Access to Ethereum smart contract state is provided by only two services: Infura and Alchemy Responses from the servers are not verified but rather trusted by the wallet NFT spec contains URL that points to the data typically a VPS running Apache Metamask under the hood simply just calls out to OpenSea to render the NFT Royalties are encoded in web2 format instead of being part of the protocol and is defeating the notion of composability Aditya Agarwal (former CTO of Dropbox) weighed and shared that while the criticism is warranted, he believes that the stateful protocols that are part of the Ethereum network are its powerful feature. Arguing that web3 protocols are different from protocols on the web developed in early 90s such as SMTP/HTTP. And the comparison is invalid. Smart contract allow data to be remixed and composed. Consider an ERC20 contract as an example. It contains attributes for Total Supply Account to Balance mapping Transfer to a Recipient for an Amount from caller’s Account Allowance for a Spender for the owner’s Account Transfer from a Sender to a Recipient for an Amount as permitted by the Allowance This contract can be utilized in other interesting ways. For instance, TokenTimelock contract allows tokens to be held with certain release times. Allowing use-cases such as “advisors get all of their tokens after 1 year”. Final words Whether the decentralization movement will actually bear any fruit is still yet to be seen. However, there is a ton of excitement in the space and lots of talented folks are looking into it. The movement has accrued interest of celebrities, politicians, academics and the every day man. There are still many paths for development and no one really knows what the future holds. But one thing we know for sure is that it is arguably the most exciting technology of the present time. I will do another post enlisting some of the various directions in which product development can occur