EigenLayer Is Unleashing the Crypto Startup Spirit

NFTs, DAOs, and DeFi have been critical forces in the gradual disruption of the financial system. Shared security is another key component of this evolution. EigenLayer has emerged as a major player in this ecosystem, and its token will become transferable on September 30.

We see this as the "AWS moment" for Web3 — a pivotal turning point, analogous to the early 2000s when server costs plummeted, as the cost of security for a crypto economy will similarly collapse. This article takes a deep dive into how Actively Validated Services (AVS) work and the reasoning behind them.

"The measure of civilization is the number of important operations we can perform without thinking about them." – Alfred North Whitehead

In late 2002, eight people attended a technology conference hosted by Amazon at their offices in the former Pacific Medical Center. Despite the small turnout, that conference marked a turning point in Amazon's destiny, the startup economy, and indeed the trajectory of capitalism itself. On that day, Amazon launched the first version of Amazon Web Services (AWS).

Jeff Barr was one of the eight attendees at the AWS launch and soon joined the AWS team. He is now AWS's Chief Evangelist and frequently blogs about new AWS features using Lego bricks. (Source)

Building an internet company in the 1990s and early 2000s required a (non-exhaustive) checklist: physical servers, networking equipment, data storage, database and operating system software licenses, a secure facility to house the hardware, teams of system administrators and network engineers, and robust backup and disaster recovery solutions. All of this would cost at least $250,000 and take months — sometimes years — to set up.

What's remarkable is that these infrastructure costs were almost entirely unrelated to the company's unique product or service. Whether you were building a pet store or a social media platform, you had to go through the same process from scratch. It was estimated that 70% of engineering time was spent building and maintaining data centers, leaving only 30% for actual business operations.

By introducing cloud computing, AWS fundamentally changed the economics of startups with its flexible pay-as-you-go model, eliminating the upfront investment of time, effort, capital, and headcount. Shifting infrastructure from capital expenditure to operating expenditure allowed small teams with revolutionary ideas to rapidly deploy and validate their hypotheses. Many of those teams eventually grew into companies like Stripe and Airbnb.

Around the same time, an anonymous programmer named Satoshi Nakamoto revolutionized the structure of capitalism in a different way. He figured out how to get globally distributed computers to reach consensus without trusting one another — solving a problem that had haunted computer scientists for decades, and representing a breakthrough moment in the history of technology.

While Satoshi's Bitcoin primarily leveraged this trustless distributed system to maintain a payment ledger, Vitalik Buterin created Ethereum, extending its capabilities to support any general computation. Over time, other use cases for this system emerged — from distributed storage networks like Filecoin to oracle networks like Chainlink that can securely deliver real-world data to blockchains.

However, building a distributed network from scratch is analogous to standing up internet companies before AWS — expensive, resource-intensive, and largely disconnected from the core problem the network is actually trying to solve. Given that many of these networks involve real money from the start, any mistakes can have catastrophic consequences.

When a problem affects enough people, solutions emerge. Amazon made it easier to start internet companies; now, the EigenLayer team is offering similar support to those who want to build trustworthy distributed computing networks. Each network built on EigenLayer is called an "Actively Validated Service" (AVS).

Before diving into AVS, we need to understand why launching a distributed network is so difficult in the first place.

The Challenge

Consider the problem: you have a network of computers spread across the globe, each running independently, and you need to achieve consensus among these untrusting nodes on a shared truth. That truth can be anything — token balances in an account, NVIDIA's stock price, the result of a complex computation, or the availability of files on a network.

Nodes in these networks may have incentives to manipulate the truth — for example, reporting a higher token balance than actually exists. However, as long as the majority of nodes agree on the actual truth, participants acting maliciously will be ignored. The situation becomes dangerous when a majority of nodes agree on a state that diverges from reality, threatening the network.

Satoshi cleverly combined concepts from cryptography and game theory to create Bitcoin's Proof-of-Work (PoW) system to address this problem. Today, most networks use variants of PoW — Proof-of-Stake (PoS) — which consists of four core components:

Cryptography: prevents impersonation and ensures the integrity and authenticity of data on the network. Reward mechanism: incentivizes legitimate participants (validators) through transaction fees from users and newly minted tokens in the network. Penalty mechanism: malicious participants face financial penalties; validators must stake the network's native token to participate, and they risk losing their staked tokens if they act maliciously (slashing). Power of distribution: having more validators, with well-distributed stake, makes the network more resistant to attacks. PoS networks allow ordinary users to delegate their tokens to validators and receive a share of their rewards. However, this approach also exposes users to risk — if the chosen validator acts maliciously, the user's stake may be slashed.

On some blockchains (e.g., ETH and Solana), protocols allow stakers to exchange their native tokens for liquid tokens (e.g., Lido provides stETH tokens to Ethereum stakers), which are derivative assets known as Liquid Staking Tokens (LSTs).

With that context in mind, imagine you're a team looking to build a PoS network from scratch. First, you need to find a set of validators — people with the technical expertise and hardware ready to join your network. You can find these individuals on Discord and X (formerly Twitter). However, to capture their attention among countless competing projects, you'll need to excel at marketing or secure significant venture capital funding.

Even if you do get their attention, convincing them to join your network is no easy feat. Remember, validators will need to stake their own capital as collateral or put in the effort to attract others to stake. Since your network is still in its early stages, the token's value may be low. Why would validators risk exposure to a token that could drop sharply at any moment, especially when they're already dealing with the volatility of other network assets?

Your best bet may be to inflate rewards: offer higher yields to validators (and stakers) to compensate for the additional risk they're taking on, which also explains why newly launched networks typically offer high staking APYs. The problem is that high issuance is essentially an indirect cost to the entire network, potentially diluting token value.

Even if you manage to overcome those hurdles, the number of validators in the early stages may still fall short of the ideal. A scarcity of validators reduces network security, leaving it vulnerable to majority attacks. On top of that, you need to consider additional factors like the geographic distribution of validators, creating secure and audited client software, and — depending on the project's specific needs — planning for infrastructure elements such as data availability, transaction ordering, confirmation services, and block proposals.

Just like internet startups before AWS, these steps are time-consuming and resource-intensive, and they're not directly related to the core problem your network is actually trying to solve.

Security as a Service

I recently explored how the internet created a new generation of businesses (platforms) that generate value by efficiently connecting supply and demand. In the scenario we just discussed, there's a group of validators — the "supply side" — looking to earn income by offering their technical services while minimizing financial risk. And the "demand side" consists of emerging blockchain protocols seeking reliable, trustworthy validators to secure their networks.

EigenLayer is a platform that bridges this gap, connecting validators (called "operators") with networks (called "Actively Validated Services" or AVS) that are seeking their services.

Now, think from the perspective of a new protocol developer:

First, EigenLayer provides a pool of vetted, trusted validators (i.e., "operators") who are committed to validating multiple services, including your emerging one. This directly solves your initial problem: how do you find reliable validators?

Second, EigenLayer's most important breakthrough is the decoupling of "rewards" from "penalties." Operators don't need to stake your native token to secure your network. EigenLayer requires them to deposit (or attract stake in) existing assets like ETH and Liquid Staking Tokens (LSTs). If malicious behavior occurs, those assets get slashed.

This separation means stakers and operators can avoid the risk of holding additional new tokens; they can earn extra yield by holding mature assets they already trust. (Saurabh's analysis of "intersubjectivity" explains how EigenLayer improves capital efficiency.)

From a protocol's perspective, this model eliminates the need to compensate validators through token issuance (which can lead to token inflation). Instead, you benefit from stronger security guarantees backed by ETH as collateral. In practice, this flexibility even allows you to opt out of issuing a token altogether, if you prefer.

Finally, you can carefully curate your operator set based on your product's specific security requirements. Before integrating them into your network, you can evaluate validators' technical capabilities, the size of their staked assets, their geographic location, and their security track record across other networks. This level of selectivity is a privilege compared to the daunting task of building a network from the ground up.

When one security risk is mitigated, another emerges — dependency on EigenLayer itself. However, EigenLayer is not a standalone blockchain; it is a set of smart contracts deployed on Ethereum. Ethereum has over 6,000 nodes and more than $86 billion backing it. While smart contract risk remains, Ethereum itself is the most secure blockchain in existence.

You might ask: "What about rewards? How does the economic model on EigenLayer work?"

Protocols can reward operators and delegators with any ERC-20 token. In practice, this gives AVS two options:

Distribute rewards in mature tokens like ETH or stablecoins. In this case, the relationship between operators and the AVS is transactional — operators provide the service, and the AVS pays them in widely accepted currency. EigenDA, the first AVS, bootstrapped operator rewards by distributing ETH to operators and delegators. Distribute rewards using their own token. This approach more closely resembles the traditional economic model of crypto networks. While this model gives AVS the flexibility to pay for security guarantees through token issuance (as opposed to the direct cost of using ETH/stablecoins), they also need to convince operators that their token will hold its value. Failing to do so will make it difficult to attract operators, as they're likely to sell AVS tokens as soon as they receive them. Initially, 10% of AVS rewards will be allocated to operators and the remainder to stakers, though this parameter will become flexible in the future. Additionally, to "strengthen incentive alignment," EigenLayer intends to distribute the equivalent of 4% of total initial $EIGEN supply as rewards, incentivizing delegators and operators to participate in the network.

EigenLayer's compelling value proposition has attracted a diverse array of projects looking to deploy as AVS. The list includes mainstream players — such as rollups that need operators, data availability services, bridges, oracle networks, and sequencing layers.

However, since operators can theoretically support any type of computation (not limited to state transitions), we're also seeing many innovative and experimental projects leveraging EigenLayer for development. These projects include decentralized physical infrastructure networks (DePIN), AI inference engines, zero-knowledge proof co-processors, privacy-focused protocols (including TEEs, FHE, and MPC), zkTLS networks, and even policy tools for smart contracts.

The Great Unleashing

Earlier, I made the relatively bold claim that AWS changed the nature of capitalism. Before AWS, the high capital requirements to start companies meant founders had to self-fund or secure investment from external sources (friends, family, venture capital). This financial barrier effectively excluded the vast majority of the global population from internet entrepreneurship, making it the exclusive domain of the wealthy or those in privileged regions.

By removing these barriers, AWS didn't just streamline the process for existing entrepreneurs — a relatively small group — it unleashed the creativity and imagination of many who previously believed entrepreneurship was out of their reach. This democratization fueled a wave of business experimentation. While many ventures failed, those that succeeded drove unprecedented economic productivity and improved quality of life for people everywhere.

From the perspective of individual entrepreneurs, cloud computing opened up a wide spectrum of possibilities — from attempting to build the next billion-dollar company to solo developers like Pieter Levels earning millions of dollars a year, and everything in between.

We're excited about EigenLayer and AVS because they unlock the same kinds of opportunities for trustless distributed networks. Got an idea that requires multiple computers to run without trusting each other? You can now rapidly bring that idea to life through an AVS.

We're already seeing a wave of experimentation — from governance chains to zkTLS networks — that may not have been feasible before. As more entrepreneurs recognize the significant reduction in human and financial capital required to build these kinds of systems, we expect to see even more experimentation.

Most will fail.

But some will light the way forward for this industry.