I remember the first time I saw a modular blockchain pitch deck. It was early 2023, in a co-working space in Nairobi, and the founder spoke with the kind of fervor that usually accompanies a religious conversion. He gestured at a diagram that looked like a stack of LEGO bricks – execution layer, settlement layer, data availability layer, consensus layer – each block a different color, each promised to be trustless, scalable, and open to all. The audience nodded, mesmerized. I sat in the back, my arms crossed, feeling the weight of five years of smart contract audits in my bones. I had seen too many beautifully drawn architectures crumble under the weight of their own assumptions.
That memory returned last week when I read about the EigenLayer restaking protocol’s latest update to its data availability committee. The update, which introduced a new slashing mechanism for validators, was celebrated across Twitter as a step toward decentralized security. But as I dug into the technical specifications, I found something unsettling: the committee’s membership was still controlled by a single multisig wallet, held by three individuals from the EigenLayer team. The code was open, but the governance was not. The modular dream was, once again, encountering the hard reality of power.
Tracing the moral code behind every token.
To understand why this matters, we need to revisit the philosophy behind modular blockchains. The core idea is simple: instead of one monolithic blockchain doing everything (execution, consensus, data storage), you split these functions into separate layers that can be optimized independently. This allows rollups to scale transaction throughput while relying on a base layer for security. Proponents argue that this is the natural evolution of blockchain design – a move from rigid, monolithic structures to flexible, composable systems. And in theory, they are right. But theory is not the same as practice.
In practice, modularity introduces new points of centralization. The most critical of these is the data availability layer. In a modular stack, the rollup does not need to download all the data itself; it only needs to verify that the data is available. This is done through data availability sampling (DAS), where light nodes randomly sample small portions of the data to confirm it has been published. The challenge is that DAS requires a massive number of light nodes to be effective, and those nodes need to be distributed across the network. If the data availability layer is operated by a small set of permissioned validators, the entire system becomes vulnerable to collusion or censorship.
Take Celestia, the first dedicated data availability layer. Its mainnet launch in 2024 was a significant milestone, but the validator set is far from decentralized. According to the latest data, over 60% of the voting power is concentrated among the top 10 validators, many of which are operated by the same venture capital firms that funded the project. This is not unique to Celestia; EigenDA, the data availability layer built on top of EigenLayer, has a similar concentration. The EigenLayer restaking mechanism was supposed to allow anyone to become a validator by staking ETH, but in practice, the barrier to entry is high. You need to run complex infrastructure, maintain high uptime, and deal with slashing risks. Most individuals cannot afford to do that, so they delegate to professional operators, leading to the same centralization we see in Ethereum itself.
Building libraries where others build empires.
I have seen this pattern before. In 2017, during my work on the ZEIP-20 standardization, I audited over 150 proposal drafts for token transfer logic. I found that many of the edge cases in the ERC-20 standard favored centralized validators by allowing them to front-run transactions. The code was neutral, but the implementation was not. The same is happening now with modular blockchains. The architecture is designed to be decentralized, but the incentives and operational realities push toward centralization. It is not malice; it is the nature of building complex systems without a corresponding focus on governance.
Consider the role of sequencers in rollups. Sequencers are responsible for ordering transactions and submitting them to the layer-1. In most rollups today, the sequencer is a single entity – either the rollup team itself or a designated operator. This creates a single point of failure. If the sequencer goes offline, the rollup stops producing blocks. If the sequencer is malicious, it can censor transactions or reorder them for profit. Decentralized sequencer sets are being researched, but they are not yet production-ready. The problem is that decentralized sequencing requires a consensus mechanism among sequencers, which reintroduces the same scaling challenges that modularity was supposed to solve.
And then there is the upgrade mechanism. Every modular rollup has a smart contract on the base layer that controls the logic of the rollup – the bridge, the state root, the withdrawal process. That contract is almost always owned by a multisig wallet, controlled by the rollup team. In theory, the team can use this multisig to upgrade the rollup, fix bugs, or respond to emergencies. In practice, the multisig is a centralized backdoor. If the team is compromised, or if the multisig signers collude, they can steal all the funds in the rollup. This is not a hypothetical risk. In 2023, the Optimism team had to use their multisig to upgrade the bridge after a vulnerability was discovered. The upgrade was necessary, but it also demonstrated that the rollup was not truly trustless.
Preserving the human story in digital ledgers.
During my work on the African AI-Blockchain Ethics Charter, I consulted with farmers, technologists, and policymakers across East Africa. One of the recurring themes was the fear of losing control to distant, opaque systems. They were not afraid of technology itself; they were afraid of the power it concentrated. The modular blockchain narrative promises empowerment, but if the underlying infrastructure is controlled by a small group of investors and developers, then the promise is hollow. I have seen this play out in the NFT space, where the OpenSea royalty surrender of 2022 destroyed the creator economy for small artists. The market celebrated efficiency, but the artists lost their livelihoods. Community over capital, always.
So what is the alternative? I do not believe we should abandon modularity. The technical benefits are real – faster settlement, lower fees, greater scalability. But we must be honest about the trade-offs. A modular blockchain is not decentralized if the data availability layer is concentrated, if the sequencer is a single point of failure, or if the upgrade mechanism is a multisig. The industry needs to develop standards for decentralization audits that go beyond counting nodes. We need to measure the distribution of power across all layers – governance, data, sequencing, and upgrades. We need to incentivize the creation of open, permissionless alternatives, such as DAS protocols that use light clients without staking requirements, or sequencer selection algorithms that are resistant to collusion.
Walking away from the hype to find the soul.
One of the most hopeful developments I have seen is the work on shared sequencer sets. Projects like Astria and Espresso are building decentralized sequencer networks that can be used by multiple rollups, reducing the reliance on a single sequencer. These networks use a proof-of-stake mechanism with a large validator set, similar to Cosmos. But they are still in early stages, and the economic incentives are not yet proven. Another approach is to use forced inclusion mechanisms, where users can bypass the sequencer and submit transactions directly to the base layer if the sequencer is censoring them. This is already implemented in some rollups, but it is slower and more expensive, making it a last resort rather than a practical solution.
From my perspective, the most critical missing piece is education. When I founded my crypto education platform in Nairobi, I realized that most developers do not understand the governance implications of the protocols they build on. They learn how to write smart contracts, but they do not learn how to audit the trust assumptions of the layers above. I have made it my mission to teach this – to show that ethics is not a feature; it is the foundation. In my courses, I spend two weeks on the concept of “trust minimization.” I ask students to map out every point where a human or a multisig can intervene in a protocol. They are often shocked to find dozens of such points. The same should be done for every modular blockchain that claims to be decentralized.
Listening to the silence between the blocks.
I recall a conversation with a young developer in Lagos who was building a DeFi lending protocol on a modular rollup. He was excited about the low fees and high throughput. I asked him who controlled the sequencer. He did not know. I asked him who owned the upgrade multisig. He said he had not checked. He was building on a platform that he assumed was trustless, but he had never verified the assumptions. That is the danger of the modular narrative – it sells a vision of the future that is not yet realized, and developers build on that vision, trusting that the details will be sorted out later. But in blockchain, the details are everything. The code is the law, and if the code contains a backdoor, the law is broken.
So what does this mean for the current bull market? The hype around modularity is at its peak. New rollups are launching every week, each promising to be the ultimate scaling solution. The market is flooded with tokens that claim to be the infrastructure for the next generation of decentralized applications. But underneath the surface, the same patterns of centralization persist. The venture capital firms that funded the modular projects are the same ones that sit on the validator sets. The multisig signers are the same people who were on the founding teams. The decentralization is a marketing claim, not a technical reality.
I do not write this to discourage innovation. I write this because I believe in the potential of blockchain to create a more equitable financial system. But that potential will only be realized if we are honest about the limitations of current designs. Community over capital, always. We need to demand transparency from the teams building this infrastructure. We need to ask for proof of decentralization, not just promises. We need to educate ourselves and others about the hidden points of control.
Ethics is not a feature; it is the foundation.
I will end with a question that I ask myself every day: If the modular blockchain is not decentralized, then what is the point of building it at all? The answer is not simple. There are use cases where speed and cost matter more than trustlessness – for example, in gaming or social media. But for DeFi, where billions of dollars are at stake, we cannot afford to compromise on decentralization. The modular architecture is a tool, not a solution. It is only as decentralized as the layers that compose it. And until we address the concentration of power in data availability, sequencing, and governance, we are building castles in the sand.
I will continue to audit, teach, and advocate for a more honest approach to blockchain design. The technology is too important to be left to the hype cycle. Tracing the moral code behind every token.
