Hi friends 👋,
I hope everyone has a great (American) Thanksgiving. No doubt this year’s turkey dinners will be dotted with conversations about the curious internet sideshow that is Web3. My wish for you all is that these articles provide frameworks that can help you survive your respective familial interrogations.
We talk a lot about the composability of software on this newsletter, but there’s something to be said about the composability of ideas. Learning compounds too. Ideas themselves are like lego blocks. And as your set of ideas increases, the number of connections you can make between them increases even faster - exponentially really.
The number of ideas for articles I have is off the charts (you should see my Notion dashboard). There’s so much to learn and thus so much to explain. The frontier moves so fast that I could spend every article chasing it. This week, I feel it’s important to address some of the open problems and questions that are well within the fold: like what will the future of Web3 applications will look like. This was a great thought exercise, I learned a ton, made some great frameworks, and let me tell you - the future looks exciting.
Let’s get to it 🚀
Emergent Apps & the Great Rebundling
The future of Web3 applications
It’s no secret that decentralized apps suffer from a poor user experience: expensive gas fees, multi-second confirmation times, an overall anxiety-inducing transaction process... I remember practically sweating through my shirt when I sent my first Ethereum transaction, knowing that one mistake in a 64-character alphanumeric code would banish my funds to the ether (pun intended). Today, I’m comfortable skittering from chain to chain, shuffling funds between burner wallets, and engaging in baseline security practices. But reaching this point was hard fought, as it continues to be for Web3 newcomers.
What is most commonly referred to as Web3’s ‘UX problem’ is actually an issue with blockchain scalability. Except at this point, many consider scalability to be a solved problem, making it out to be a huge red herring. In fact, our current scalability-constrained Web3 experience has done nothing but thrive! On-chain activity continues to grow in spite of a poor experience. NFTs, DAOs and DeFi have proven enticing enough to overcome UX hurdles that would crush even the greatest SaaS product. Combine this with a known roadmap for rolling out L2 solutions, and it feels like we should be ready to move on. That’s why it’s time to shift the conversation to the UX problem of tomorrow.
Today, a Web3 user’s experience is spread out across an ever-growing list of decentralized apps and websites. You can buy NFTs on Opensea or enter into a group-bid on PartyBid. DAO contributors can find open tasks on a Notion bounty board and submit rewards for contributors on Coordinape. Communities plan their initiatives on Discord and vote on proposals using Snapshot. Web3’s product velocity has set off an explosion of new protocols and helpful services, but in doing so we’ve fragmented our experience, making the space more intimidating to participate in.
We must remember that Web2 apps are highly adapted to the design principles of user needs. The experience we have using apps and browsers today survived decades of iteration, and are likely continue to dominate the on-screen experience. That’s why I’m predicting that a Great Rebundling will occur - that many of today’s Web3 services are destined to become embedded features rather than explicit experiences.
Remember that protocols are most powerful when they are widely used, universally composed, and built upon. That’s why the natural position of many of Web3 products will be to act as integrations rather than provide a full-stack experience. To fully engage in this discussion, we’ll need to imagine the future of Web3 UX, building out a picture for the first applications that can rightfully call themselves ‘Web3 native’.
A World of Fragmented UX
Protocols as Behavioural Coordinators
Hypertokens: Embedded Token Experiences
Composable Services, Emergent Apps
A World of Fragmented UX
You’ve likely noticed that software development roles tend to be split into three categories: front-end, back-end, and full-stack. This is representative of the three main layers to a modern application:
A front-end for the user to interact with
A back-end to store all the data
Some connecting logic and integrations that give user actions the ability to edit and manipulate data
Today, building an application requires owning each layer of this stack, whether directly by hiring designers to create your front-end, or indirectly by renting databases from providers like Azure or AWS. We converged on this model for app architecture after years of iteration, but building a Web3 app requires that a new model be conceived from the ground up.
Those who have been along for the Future Proof journey know that Web3 is defined by the rise of open state networks. These networks are similar to the internet. They are networks of computers working together to uphold a digital experience. Except unlike the internet, open state networks use blockchains as a shared dataset that anyone can access. A shared dataset sounds dangerous, which is why cryptography-enabled protocols ensuring that only those with the right access can manipulate sensitive data, like your Bitcoin wallet.
The implications of having an internet that stores our data are vast. It changes the very nature of what apps can do and how they are built. Right away we can imagine that programs built on Web3 no longer need their own database. The network becomes their database. What’s more is that much of the logic needed to read and write to the network are standardized into open protocols that anyone can use. So we can see that much of an app’s ‘job’ is already being accomplished by the network itself.
So it appears that Web3 apps are restricted to competing on the uppermost interface layer and some of the integration layer. Danny Zuckerman, co-founder of Ceramic Network, describes this new app architecture model in his article on data composability (which I highly recommend you read!).
Consider a Web3 equivalent to Medium or Substack. These apps are expected to own the user interface, but no longer need to own their full stack. Danny imagines a world of shared services that provide apps with the logic needed to create a writing editor or a subscription feature. These services will manage any data stored on the network - like your unpublished drafts or a list of subscribers. And because both data and services are available for all to use, Medium and Substack don’t need to build these components from scratch. Instead, they can leverage the common services and specialize in their unique value proposition.

This framework speaks to the principle of Web3 composability, that we will have highly interoperable snippets of code that can be bundled together into any app experience. It’s a bold future where apps and their interfaces are enabled to become highly tailored to user needs, yet this is not what we’re seeing in the ecosystem.
Today, most Web3 services/protocols vertically integrate their user experience. In other words, they own both the service and the interface. A few examples:
Yup.io is a Web3 protocol for content curation. Yup offers a browser extension that allows users to rate and categorize content through the right-click menu (service). Yup also owns a website that displays user profiles alongside their curation content (interface).

Coordinape is a protocol for rewarding project contributors in proportion to the value of their work. Users go to Coordinape to allocate their points across the team, which allows payment to reflect the collectively recognized impact of everyone’s effort (service). Coordinape also hosts several web pages for visualization and voting (interface).
PartyBid allows users to create and join group-bids on NFTs (service). They own a search experience where you can scroll through existing group-bids, or create a bid of your own on their website (interface).
Recall my claim at the top of this article: that most Web3 services are destined to become embedded features instead of explicit experiences. Content curation, collective voting, group-bidding - these feel more like features waiting to be bundled than apps themselves. And that’s alright! Being bundled is actually good for protocols, as it increases their reach and stature at no cost to the team. We should expect apps to integrate protocols rather than protocols building out apps. However, Web3 waits for nobody, and as a result we see protocol teams bootstrapping their own UX, where only some will make the leap to being full fledged apps.
A great example of where this is happening today is with Mirror’s crowdfunding campaigns. Mirror, originally a content platform similar to Medium, has bundled together features to enable collective funding. A recent example, KrauseHouse, a DAO with the sole intention of purchasing a basketball team, raised $1.5M through their crowdfund campaign by offering ‘Genesis Ticket’ NFTs and early governance tokens for varying levels of contribution. Both the NFTs and the ‘buy now’ button were embedded in the same page of content that detailed their planned roadmap to NBA-ownership.

This is the type of experience we should expect to see going forward. Yes, an NFT in isolation can represent art or a digital collectible, but it is most useful when it is seen as an embedded feature of the internet. There is great nuance in the difference between buying a Genesis Ticket on Opensea and purchasing it in-context through KrauseHouse’s fundraising article.
History provides us a good example of this embedding in action. The early internet began as a purely text-based experience. Web pages were just text files hyperlinked together to enable page-to-page navigation. But as the internet became more dimensional, it added multimedia experiences like video, audio, and image data. These new features of the internet didn’t remain in isolation on video-only platforms, or image-only applications - they were added to the internet mesh and we went from hypertext to hypermedia.

In other words, the internet became more dimensional. The web page stayed the same, but it could do and be more, at least within the confines of multimedia. The creation of Web3-enabled apps will mirror this entanglement. We should expect protocols and tokens to join the mesh, revitalizing the Web2 experiences we know and love now the with new properties enabled by Web3. We should think of Web3 as becoming hypertoken.
This is a helpful mental model, but the framework is still a blank canvas. We want to know: What will these token-embedded experiences will look like? What will it feel like to use a Web3 app? If we want answering these questions to be more than just speculation, we’ll need to take it back to principles. That means getting incredibly precise about the unique capabilities of what Web3 offers. And if there is any place to start, it’s with Web3’s dynamic duo: protocols and tokens.
Protocols as Behavioural Coordinators
We know broadly that Web3 is the idea that open stateful protocols will dominate the internet experience. So what will it look like for protocols to join the internet mesh?
In Meta vs. the World, we learned that protocols facilitate an exchange of value between consumers and suppliers. We tend to think of this as a network of computers getting paid in return for compute services or memory storage. But this just happens to be the form factor of the protocols we have today.
Protocols are actually highly versatile tools that can be used to coordinate the supply for any type of work. This is best emphasized by one of my favorite Chris Dixon quotes:
“Software is simply the encoding of human thought, and as such has an almost unbounded design space.”
We already know what can be accomplished when protocols organize the work of computers (Bitcoin, Ethereum, Arweave, etc.). A far more interesting application of protocols is when we view humans as suppliers. An example of this is Yup.io’s content curation protocol:
Yup incentivizes users to create lists of great articles and videos, and to rate creators based on the quality of their content. When a user recommends content that is later strongly endorsed by the community, they receive $YUP tokens as a reward for their good taste. The end result is that curators are motivated to aggregate good info and identify top creators. Seeing as we’re all attention-limited, this seems a great feature to embed across many websites. Indeed, the highest calling of Yup is to become the like/dislike/react feed for the entire internet!
This is a powerful example for how protocols can become a type of behavioural coordinator, giving users an incentive to create a desired supply - in this case, a supply of good content. There are many valuable services that require human judgement as a production input, and protocols can coordinate supply for them all.
The dynamic of behavioural coordination will undoubtedly be present in Web3 apps. Protocols will find ways to align human effort, creating a market for services unlikely to exist in Web2 solutions.
Imagine Wordpress, which is already an open source platform, except with a protocol coordinating the creation of new UI components. Developers can be rewarded when their components are used in websites, and are incentivized to build solutions for unmet user needs. Picture how this might changes networking, if a protocol exists that encourages peers to accurately label each others’ skills and experiences. Such protocols only need to be built once (correctly!) to be embedded into apps across the entire internet.
Hypertokens: Embedded Token Experiences
Anyone who has participated in Web3 knows that the user experience is mainly centered around the transaction. Any time a token needs to move between wallets or contracts, a user must be in the loop to provide their cryptographic signature. This experience isn’t inherent to tokens, but is actually a byproduct of the context we put tokens in, namely as currencies and online goods. Chris Dixon famously challenged this worldview in, Tokens Are a New Digital Primitive, saying:
“There is nothing intrinsic about fungible tokens that needs to be related to money and finance, and nothing intrinsic about NFTs that needs to be related to art and collectibles.”
Indeed, using tokens as currencies and collectibles is another example of Web3 skeuomorphism - the tendency we have of bringing old design principles from the physical world to new use cases in the digital world. In reality, tokens inherit software’s unbounded design space, and can be made to do anything.
I believe that our view of Web3 as being a transaction-based experience will fade as tokens and protocols join the internet mesh. Similar to images and videos, we can picture an enriched internet, with tokens layering exciting new functionality on top of the humble web page.
Adding embedded token experiences to the mesh is the greatest leap in dimensionality the internet has seen yet. Each token can connect to a new network or protocol, imbued with its own incentives. Because the design space is so large, instead of examples we’ll approach this from principles. Below, I’ve distilled five essential token properties that are likely to become part of our app and internet experience.
Financialize: Tokens give all goods and services an explicit financial value by connecting them to a global market. This includes denominating the value of society’s intangibles, allowing creative work (which often derives value from culture and influence) to find a true price in markets.
Incentivize: Tokens are the mechanism of influence for an incentive system or protocol. They align network participants by rewarding them for positive-sum behaviour (i.e. curating good content), and creating barriers for negative-sum behaviour (i.e. submitting false transactions to the Bitcoin network).
Fractionalize: Tokens increase the resolution of ownership, making economic participation more accessible. It also becomes easier to co-own assets, that would otherwise be difficult to share.
Collectivize: Tokens catalyze community formation around narrative and shared purpose. On-chain commitments act as social proof that can launch communities past the chasm from ideation → formation. ConstitutionDAO’s strong narrative brought thousands of users under the same banner with tens of millions in funding.
Authorize: Tokens are a natural way to grant access to online experiences. Friends with Benefits DAO limits access to their Discord, newsletter, and even IRL events based on the quantity of tokens you hold. Similarly, any creator or developer can release a token-gated experience that unlocks content or app features based on a user’s buy in.
These properties have the potential to transform all our online experiences. Imagine Zillow with the ability to group-bid on a house (fractionalize), or a social EdTech platform that creates study groups based on your learning performance and history (authorize). The intersection of these properties are even more fascinating: Picture a municipality-run app where communities vote on crowdsourced proposals for infrastructure development (collectivize). The city can ‘buy’ the winning proposals from their creators (financialize), implicating the neighbourhood in the financial success of the project (incentivize).
Yes, some of these experiences can be created with the software principles of Web2, but that would negate the real benefits here. Anyone can access Web3 services! Your city can choose from a bundle of services to create their proposal-bidding app, instead of using taxpayer money to develop an in-house, full-stack solution. Existing apps like Zillow or Coursera can provide quick value to their users by integrating, rather than building, new features. There is immense value in permissionless services, but to realize it we must accept that apps are more composable than originally thought.
Composable Services, Emergent Apps
Composable apps are a radical new paradigm that changes how value is created in the world of apps. Now, instead of building out full solutions, developers will look to either build new services or bundle new services underneath compelling interfaces.
Today, we are at an in-between stage. There isn’t enough activity at the interface layer to support pure play services, and there are also huge gaps in both the services and interface market that are highly lucrative to fill. As a result, out of necessity and ambition, we see many services expanding their presence into the interface layer.
This is indicative of our early stage in the Web3 app ecosystem. In the beginning, it’s a land grab. Everyone is racing to become a ‘canonical service’ - a culturally accepted protocol for their given use case. Such services can embed themselves as a de facto layer in the value chain, receiving the lion’s share of fees in their market. But as the space of services fill out, there will be greater incentive to specialize services and meet user needs off the beaten path.
This will lead to a more diverse library of services that can be bundled into experiences. Somewhere along this journey, competition will move to the interface layer. Here, a majority of dev time will be spent creating experiences that consumers love, enabled by the most widely used protocols and services from Web3.
We can imagine how this makes existing apps more dynamic:
Web3 Twitter can fill its feed with content selected by the best content curation protocol, and at the same time, offer creators the ability to get paid by token-gating a portion of their content. For users, it will be the same Twitter experience, with the added opportunity to enrich their feed by buying into their favorite creator’s tokens to unlock top content. I also imagine a world where different content curation protocols exist for different purposes: one for learning, one for entertainment, and one for your family and friends. These could all be developed separately and integrated into any app, alternated between by the interface layer.
Web3 Patreon can offer a subscription service for exclusive creator content, where content is token-gated using the same service as Web3 Twitter. Web3 Patreon can integrate audience engagement features, allowing audiences to vote on new ideas for content - either sourced from the community or proposed by the creator. Each patron’s vote is weighted in proportion to their holdings of the creator’s token. The community collectively pools their tokens to decide the content they want to the creator to make next.

Composability allows an app to be greater than the sum of its parts, which is why apps will become emergent. Users will receive interfaces with niche-specific functionality that more precisely meet their needs, and builders will achieve greater product velocity and faster time-to-market by leveraging the work of others.
Embracing this new world of app development requires retraining how we think about app-competitiveness. It’s no longer about who can write the most code, and it’s certainly not about who can hoard the most data. This is a more even playing field, where apps succeed by the virtue of meeting user needs and delivering great experiences. This wave will raise all of society up with technology that is permissionless, transparent, and accessible to all.
How exciting the future will be.
If you enjoyed this article consider subscribing for weekly content on the technology that will shape our future!
If you really enjoyed this article consider sharing it with an intellectually curious friend or family member!
Let me know what you thought about this piece on Twitter.
With gratitude, ✌️
Cooper