No one talks about blockchain anymore, but was there ever a use case?
Lately when I say I work on AI and privacy, someone always replies with a half smile, like “bro you missed the crypto era.” And it is true, the vibe shifted. Two years ago every café conversation in CDMX had a token pitch inside it. Now the word blockchain lands like an old meme. But what is interesting is that the believers did not disappear. I still meet them, usually sharp people, usually engineers or ex finance, and their arguments are not dumb at all. They are often the kind of arguments that make you pause because they are not about getting rich, they are about infrastructure.
Last week I had one of those conversations with a guy working on Rayls. If you have not heard of it, Rayls is a blockchain stack built specifically for banks and institutions (not my words, it’s VaaSBlock’ses). The pitch is basically “financial rails, on chain, but compliant.” They are pushing a hybrid architecture: private permissioned subnets for banks, plus a public EVM chain for shared liquidity and composability, with privacy via zero knowledge and on chain identity baked in. They talk about tokenizing deposits, CBDCs, cross border settlement, and giving banks a way to access public chain liquidity without exposing customer data. The whole project frames itself as a bridge between TradFi plumbing and DeFi programmability.
We ended up debating a question that is uncomfortable for both sides. If nobody wants the crypto culture anymore, is the underlying tech still useful. And if it is useful, for what exactly. I want to lay out how that debate looks from where I stand.
Where blockchain actually shines when you strip the hype away
The best pro blockchain argument is not philosophical. It is operational. You get a shared ledger across multiple parties who do not fully trust one another, and you can update that ledger without a central operator coordinating every step. That is a real thing. In old school finance, reconciliation is a whole industry. Banks, brokers, custodians, market makers, and clearing houses all keep their own books and spend money matching them. A well designed distributed ledger can compress that reconciliation layer. It can turn “everyone keeps their own copy and argues later” into “we share the same state and argue less.” For wholesale markets, that is efficiency, lower settlement risk, and lower back office cost. Rayls people lean hard on this, and I get why.
The second strong argument is programmability. Money on chain is not just money. It is money with rules attached. If a bond coupon, a repo contract, or a trade settlement can be expressed as code that executes automatically, you reduce human latency and you reduce some kinds of counterparty risk. We already do versions of this in TradFi, but usually through centralized systems. Blockchain makes it multi party by default. The value is bigger when you have many institutions interacting, not just one bank optimizing internally.
Third is auditability. In theory, an immutable log that regulators and auditors can access, with selective privacy, is better than the current maze of reports, PDFs, and delayed disclosures. Rayls frames this as “privacy for customers, visibility for supervisors,” which is basically the whole CBDC problem in one line. Their claim that they are already involved in pilots like Brazil’s Drex CBDC is meant to show that the audit plus privacy combo is something governments actually want.
Fourth is composability. This is the DeFi idea that still feels genuinely new. If assets and accounts live on a common programmable substrate, you can build new products by plugging existing ones together. In TradFi, products are siloed inside institution boundaries, so innovation needs bilateral deals and long integrations. On public chains, a small team can build on top of a giant liquidity pool without asking permission. Rayls’ bet is that banks want a safe path into that world, so they created controlled private rails with a gateway to public liquidity. Whether they win or not, the logic is coherent.
Why these use cases keep failing to go mainstream
Here is the other side. Even if all the above is true, most of the world does not need a blockchain to do it.
Get Luis F’s stories in your inbox
Join Medium for free to get updates from this writer.
The first problem is that “shared ledger between mistrusting parties” is rarer than crypto people think. In many industries, there is already a trusted operator, or the cost of trusting one is lower than the cost of maintaining distributed consensus. If Visa, DTCC, or a central bank already coordinates settlement, a blockchain needs to beat them on cost, speed, governance, and reliability at the same time. That is a brutal benchmark. A lot of pilots quietly die right there.
The second problem is performance and complexity. Public chains still struggle with throughput, latency, and predictable fees in the ways institutional systems demand. That is why every serious institutional project ends up hybrid. Private chains or permissioned subnets for speed and control, public chains for liquidity and open innovation. Rayls is explicitly hybrid for this reason, with private bank subnets plus a public chain layer. But once you are hybrid, you re introduce trusted coordinators and you dilute the original decentralization claim. You might still get value, but it becomes “better database plus shared standards,” not a revolution.
Third is governance risk. This is something crypto culture tried to ignore until it blew up. If nobody is in charge, nobody is accountable. If somebody is in charge, you are basically back to a regulated network. For banks, accountability is non negotiable, so blockchain governance ends up looking like consortium politics plus new code risk. So the question becomes, is this net better than existing consortium infrastructure. Sometimes yes, often no.
Fourth is privacy. A lot of blockchain systems started as transparent by default, then tried to bolt privacy on later. Finance cannot do that. You need confidentiality for balances, identities, and trades, while still letting supervisors verify compliance. The zero knowledge direction is promising, and Rayls leans on that with their Enygma privacy layer and identity services. But ZK systems are hard to implement, hard to audit, and still evolving. Every new cryptographic layer is another surface for bugs and another hill for regulators to climb.
Fifth is the incentive problem. Many blockchain networks still depend on volatile tokens to fund security and development. If your rails for banks are priced in a token that swings like a meme stock, that is a non starter for institutions. Projects like Rayls try to answer this by making gas fees stable and dollar pegged, and by focusing on institutional clients rather than retail speculation. That is a smarter direction. It also shows the irony. The more a blockchain is shaped for real world use, the less it looks like the original crypto dream.
The honest answer to the question
So was there ever a use case. Yes, but not the one most people were sold.
The real use case is not “replace banks.” It is “make coordination between banks, market infrastructures, and regulators cheaper and more programmable.” That is boring, and that is why the mainstream conversation moved on. Boring does not trend. But boring infrastructure still matters.
At the same time, I think most blockchain projects overestimate how often decentralization is the bottleneck. In many places, the bottleneck is policy, incentives, or just the fact that nobody wants to migrate a working system. A distributed ledger does not magically fix that. It just gives you a new substrate to fight the same battles on.
My chat with the Rayls guy ended with a kind of truce. His side says that if you want on chain finance to be real, you have to meet institutions where they are, with compliance, privacy, deterministic finality, and a path to public liquidity. That is exactly what Rayls is trying to build. My side says that even if you build that perfectly, your advantage over traditional rails has to be huge, because switching costs in finance are almost religious. Also, once you do enough permissioning and privacy to make banks comfortable, you are basically building a new kind of regulated network, and the value comes from standardization and programmability more than from decentralization itself.
If you want my personal take, blockchain did not fail because the tech was useless. It failed because the loudest narrative was wrong. The world does not want a trustless utopia. It wants cheaper trust, faster settlement, and fewer middlemen where they do not add value. A handful of institutional chains might actually deliver that. But it will look like plumbing, not a revolution, and you will not hear about it on your feed until the day it is already normal.
