One Chain to Rule them All?

There are dozens of credible active layer 1 chains, with more being developed. So declaring victory for any one of them seems wildly premature. But there is another question that one can discuss more readily. When the protocol innovation sprint is over and the dust starts to settle, how many L1s will remain? Will there be one chain to rule them all?

I highly doubt that outcome because there are now – and for reasons of mathematics (i.e. not something that can be circumvented by some better idea) – will always be tradeoffs between security, decentralization, speed and other features of a chain that one might care about (eg complexity of executable contracts). Conversely, however, I also am skeptical of any outcome that has more than two dozen important L1 chains.

Why is that? Because protocols have strong network effects in a decentralized world. More projects using a protocol means better security, better tooling, better economics (assuming the right tokenomics), etc. Even in a more centralized world there are few successful protocols in any one area of computation. I am old enough to remember ARCNET and Token Ring, two alternative local area networking protocols. Today pretty much everything runs on Ethernet instead. I also remember Gopher and WAIS, today it’s all HTTP. There are many more examples like that including email protocols and even cabling standards.

So for any one area of decentralization, whether it is file storage or messaging, or smart contract execution, I find it hard to envision a world where more than a couple of protocols (and often only a single one) will really matter. If you allow for say a dozen primitives, from which everything will be built, then there is room for at most say twenty-ish L1 protocols and I wouldn’t be at all surprised if the number settled out even lower than that.