tl;dr: you can’t build Web3 with a Web2 mental model

I had the privilege of working with the Dapper Labs team, the creators of Crypto Kitties, NBA Top Shot, and the Flow blockchain, during a 6 month consulting arrangement a few years ago.

My assignment was to help write version 1 of what eventually became the Flow blockchain whitepaper.

Dapper’s may not have been the pure inventor of the concept of the NFT, but they were certainly the first ones to deploy it at scale, via Crypto Kitties, on Ethereum.

So, why, then did they feel the need to invent an entirely new blockchain?

Composability.

Composability is what they called the “superpower” of crypto/Web3 and Ethereum’s approach to scaling via roll-ups and sharding would destroy that superpower.

So, what is composability?

It’s basically standardization of interoperability between units.

The example Dapper kept using was “lego bricks” or, in the parlance of DeFi, “money legos.”

The idea that any service could use any other service in any way that it wanted to.

So, a “money lego” might be Amazon’s 1-click purchasing functionality which essentially says

“Pay for everything in this account’s shopping cart, charge their credit card, verify that its accurate, confirm the order and ship to the desired destination.”

Now, imagine if ANYONE could just take that and use it themselves, without having to rebuild the functionality themselves (since Amazon owns a patent on it).

In Web2, you can’t do that (legally). In Web3, you can and the fact that everything is open and reusable, assuming its located within the same environment (i.e. same blockchain or shard), means that a LOT of development time is cut, thereby allowing for more rapid development of highly valuable services.

However, if those legos are not located in the same location, then composability is broken.

That’s what happens in Ethereum v2, its rollups, and other sharded solutions (to the best of my knowledge).

All of this comes to mind because I was read Moving Past Smart Contracts: Natively Composable DLT with Radix and it makes a super-compelling case, in the same way that the Flow guys did, that composability is what matters.

As they write:

True composability makes DLT a powerful next-generation integration technology.

Yes, transaction speed, security, decentralization all matter, the so-called ‘trilemma,’ but in reality those are just table stakes.

For developers to create next generation DeFi, those that have access to highly composable platforms are going to have a built-in advantage over those that don’t.

However, to make that a reality requires a new language, a new mode of thinking about how technologies work together.

Web 2 was about a communications revolution. Web 3 is about a financial revolution. Smart contracts today (at least as envisioned and executed on Ethereum) are messages that send updates about ledger state, that is to say, it’s a Web2 paradigm.

The Web is a technology of communication that needed a language of messages. DLT is a technology of transactions that requires a new language of assets to make interoperability between DLT applications native, easy, and intuitive

It’s very tempting to look at the crypto industry and say “oh, it’s Ethereum and the EVM chains because of the TVL and number of developers there” and declare the financial revolution under way.

And that might be the case, but if Ethereum and EVM are the path to the future, I’m not sure that DeFi can really deliver on its full potential. It will be better, but not revolutionary.

On the other hand, if we recognize that we’re still early in this game and that there are 25 million developers who AREN’T yet on Web3 (compared to the 18k that are), perhaps there’s a lot of financial innovation still to come.

And the language they will all be speaking may not yet be obvious.