The Web3 ecosystem is often regarded as the next infrastructure of the internet. However, nearly 10 years after the release of the Ethereum white paper, we have very few mainstream applications running on that infrastructure. Meanwhile, we continue to see the emergence of new infrastructure building blocks everywhere: L1, L2, and L3 blockchains, rollups, ZK layers, DeFi protocols, and many others. While we might be building the future of the internet with Web3, there is little doubt that we are overbuilding the infrastructure layer. Currently, the ratio between infrastructure and applications in Web3 has no parallels in the history of technology markets.
Why is this happening? Simply because it’s profitable to build infrastructure in Web3.
Web3 defies some of the conventional market adoption patterns in tech infrastructure, creating both a rapid path to profitability and unique risks for its evolution. To explore this thesis further, we must understand how value is typically created in infrastructure technology trends, how Web3 diverges from this norm, and the risks posed by overbuilding infrastructure.
Traditionally, value creation in tech markets fluctuates between the infrastructure and application layers, finding a dynamic balance between the two.
Take the Web1 era as an example. Companies such as Cisco, IBM, and Sun Microsystems powered the infrastructure layer of the internet. But, even during those early days, applications like Netscape and AOL emerged to capture significant value. The Web2 era was driven by cloud infrastructure, which then triggered SaaS and social platforms, catalyzing the creation of new cloud infrastructure.
More recently, trends like generative AI began as an infrastructure play with model builders, but applications such as ChatGPT, NotebookLM, and Perplexity quickly captured momentum. This, in turn, drove the creation of new infrastructure to support a new generation of AI applications — a cycle that is likely to continue for several iterations.
This constant value-creation balance between application and infrastructure layers has been a hallmark of technology markets, making Web3 a notable anomaly. But why is this imbalance so evident in Web3?
The main difference between Web3 and its predecessors is the rapid path to capital formation and liquidity in infrastructure projects. In Web3, infrastructure projects typically launch tokens that become tradable on exchanges, providing substantial liquidity for investors, teams, and communities. This contrasts with traditional markets, where investor liquidity is typically realized through company acquisitions or public offerings, both of which usually take considerable time. Most venture capital firms operate on a ten-year investment cycle or longer. While rapid capital formation is one of the benefits of Web3, it often misaligns team incentives, discouraging long-term value creation.
This “infrastructure casino” is a risky pattern in Web3 because it incentivizes builders and investors to prioritize infrastructure projects over applications. After all, who needs applications when L2 tokens can achieve multibillion-dollar valuations in just a few years with minimal usage? This approach presents several challenges, and many of them are subtle and difficult to address.
1) Building Without Adoption Feedback
Perhaps the most significant risk of overbuilding infrastructure in Web3 is the lack of market feedback from applications built on top of that infrastructure. Applications represent the ultimate expression of consumer and business use cases and regularly guide new use cases in infrastructure. Without application feedback, Web3 risks building infrastructure for “imaginary” use cases that are disconnected from market reality.
2) Extreme Liquidity Fragmentation
The launch of new Web3 infrastructure ecosystems is one of the main contributors to liquidity fragmentation in the space. New blockchains often require billions of dollars to bootstrap liquidity and attract tier 1 DeFi projects to their ecosystems. Over the past few months, the creation of new L1 and L2 blockchains has outpaced the influx of new capital into the market. As a result, capital in Web3 is more fragmented than ever, creating significant adoption challenges.
3) Inevitable, Increasing Complexity
Have you tried using some of the wallets, dApps, and bridges for newer blockchains? The user experience is typically difficult. Tech infrastructure naturally grows more complex and sophisticated over time. Applications built on that infrastructure typically abstract away this complexity for end users. However, in Web3 — where there is a lack of application development — users are left to interact with increasingly complex blockchains, leading to friction in adoption.
4) Limited Developer Communities
If Web3 infrastructure has outpaced capital formation, then the challenge is even greater when it comes to developer communities. dApps are built by developers, and creating new developer communities is always a challenge. Most new Web3 infrastructure projects operate with very limited developer communities because they pull talent from the same existing pool, which is simply not large enough to sustain the vast amount of infrastructure being built.
5) Widening Gap with Web2
A side effect of overbuilding infrastructure in Web3 — without app adoption — is the widening adoption gap with Web2. Trends such as generative AI are powering a new generation of Web2 apps and redefining sectors like SaaS and mobile. Instead of tapping into this momentum, the predominant trend in Web3 continues to be building more blockchains.
Launching L1 and L2 blockchains is a profitable business for investors and development teams, but that doesn’t necessarily translate into long-term benefits for the Web3 ecosystem. Web3 is still in its early stages, and while more infrastructure building blocks are certainly needed, most of the industry is currently building infrastructure without market feedback.
That market feedback typically comes from applications using the infrastructure — but such applications are largely absent in Web3. Most usage of Web3 infrastructure comes from other Web3 infrastructure projects. We continue to build infrastructure, launch tokens, and raise capital, but we are effectively flying blind.
