Ethereum is eating all the blockchains. That’s fine.

If history is a prologue, then Ethereum is going to eat the entire blockchain sector and everything that is not Ethereum will eventually become an Ethereum Layer 2. I believe the recent decision by CELO stakeholders to shift into operating as an Ethereum Layer 2 is just the beginning of an avalanche of similar integrations and shifts that lead us to an end-state where Ethereum ultimately operates as the Layer 1 for all blockchains.

Paul Brody is EY’s global blockchain leader and a CoinDesk columnist.

There is ample precedent for this kind of consolidation in the technology industry, and one of my favorite examples is how an immensely diverse world of networks slowly but surely converged upon a single global standard over a period of about 15 years.

The networking story goes like this: A long time ago, roughly around the dawn of human civilization (in the 1970s), we had lots of different data networks. All kinds of them for a whole range of companies and governments from the U.S. Advanced Research Projects Agency Network (ARPANET), the precursor of the internet, to IBM’s Systems Network Architecture (SNA), Xerox’s Internetwork Datagram Protocol (IDP) and several others. The result was a veritable alphabet soup of incompatible networks that made connecting business and government systems extremely difficult.

Starting in the 1970s, a concerted effort was made to create a protocol that could work across multiple networks and smoothly handle interruptions and changes in network operation. The result was TCP/IP, which stood for Transmission Control Protocol/Internet Protocol. In the early days, TCP/IP did exactly what it was supposed to do: connect all these different networks.

At first, TCP/IP was just supposed to connect different networking standards, a job it did very well. Over time, however, the inexorable logic of standardization and scale turned TCP/IP from connectivity glue into a global standard. IP Networks ate up the networking business and today, there are virtually no non-IP networks remaining.

Read more: Paul Brody – Decentralization Is the Point, and We’re Not Talking Enough About Why

Given how much the technology industry loves scaling around a standard, nobody should be surprised by this and, similarly, we should not be surprised if the same happens to blockchain networks. Since the value of any network grows with interconnection, this approach could very well become a lifeline for struggling Layer 1s that, only very recently, were touting themselves as “Ethereum Killers.”

Not all Layer 2s and sidechains are alike, and lately I have been thinking about what are the different ways in which this Layer 2 ecosystem could develop. There are a number of highly specialized sub-ecosystems that could emerge. For example at EY, we are targeting industrial companies as users of our OpsChain solutions to help manage inventory and track carbon emissions. When we sit down and do scaling planning, the volumes we talk about are very large. For example, one of our clients is asking us to think about handling 500,000 units a day (all unique and serialized) for a single product line.

With those 500,000 units every day moving an average of three to four times between production and end consumption, we can think about an average of 2 million NFT transactions each day for a single product line. For these kinds of clients, the top priorities are privacy (keeping your detailed business operations data a secret from your competition) and scalability – they need reliably high throughput and low transaction costs. It will not surprise you that Nightfall, a Layer 2 network developed by EY and contributed into the public domain, is designed to do just that.

Financial transactions will have very different Layer 2 requirements. Some, like swaps, may just be looking for very high volume, low-cost roll-ups, while complex Decentralized Finance (DeFi) smart contracts will also need networks that support full Ethereum Virtual Machine (EVM) compatibility so the smart contracts can run on the blockchain.

Nor will I be surprised to see the emergence of highly specialized national, regional, or verified identity networks where all the participants are not only known, but all are identified and subject to the same regulatory rules. Imagine a Layer 2 only open to Accredited Investors who are U.S. “persons” (citizens or residents). That would allow all these people to transact a very wide variety of assets with each other with a minimum of added verification checks. They could quickly emerge inside the EU or in other major jurisdictions.

With all these specialized networks emerging, you might be tempted to wonder if there is any point in interconnecting them all through Ethereum. The value of interconnection, beyond mere EVM compatibility, is the ability to flow products and services from one ecosystem to another. No real, modern economic system is truly isolated. Every commercial contract closes with payment and financial services, of one kind or another, underpin all of them. Financial flows between countries and ecosystems underpin all trade and investment.

Nor is it likely we can ever build a single network that can support all the different kinds of transactions and the volume that will serve the entire world. As a result, there will always be multiple networks and there will be friction involved in connections between networks, even when it is just between a Layer 1 and a Layer 2. Even so, having Ethereum as a binding Layer 1 across many specialized networks will offer enormous benefits.

For example, industrial product tokens can leave a specialty manufacturing network in exchange for payment that comes from a finance-focused Layer 2, but the ability to have a continuous digital record across two Layer 2 networks and connected by Ethereum as the Layer 1 is an order of magnitude more integrated than anything that exists in the commercial world today.

One downside of Ethereum eating the entire world is that similar to the networking industry today, there will be much less variability in some of the network functions that are available. To be interoperable, tokens and smart contracts will basically have to be the same everywhere. Every chain has to be an EVM chain. And while you can have cross-chain development systems that work on a diverse ecosystem, it is not very useful because your tokens and smart contracts become stranded and useless, and the unique and special functions of particular networks never actually get used.

One of the big lessons from the world of technology is that, over and over again, universal infrastructure is more successful than specialized infrastructure, even when the specialized infrastructure is actually better for a specific job. Before TCP/IP ate the entire world of networks, there used to be specialized networks just for voice calls. They were called circuit-switched networks and they guaranteed the quality of your phone call. No delays, no breaks, no lost packets, just a continuously connected circuit between two phones. By comparison, Voice Over IP phones calls are a huge step backwards in quality, but they now represent more than 99% of all phone calls anyway.

So, make your peace and say your goodbyes to the cool specialized blockchains we live with today. I’m betting they’ll be history soon enough.