Op-Ed: L2s Deserve Warning Shots Fired in Bankless Episode

Blockchain tribalism often makes us deaf to our critics. We ignore them at our peril.

Despite making me squirm at points, I listened carefully to Ethereum researcher Max Resnick’s much-discussed Bankless interview in which he lambasted Ethereum Layer 2s (L2s) as “parasitic.”

Resnick's argument is that L2s are becoming a cozy cottage industry that sucks value away from Ethereum, the Layer 1 blockchain. 

The concern that L2s are problematic because they step on the toes of L1, attracting similar use cases, feels like upside-down thinking to me. Blockchain does not exist for the glory of any particular chain or project. L1s cannot cope with the mass-use that all of us in this space expect to see. 

And it’s a fact that L2s are the most efficient, most ethos-aligned solution to this. They are a solution for devs, a solution for users. And actually, in the long-run, they enable L1 to have more relevance to more people — not less — as more people will be using systems anchored by its security. 

Going deeper, Max views companies behind those L2s as dragging out the time it takes to decentralize the blockchain because of the high revenue fees being generated through the centralized sequencer — fees they would need to share if the sequencer was decentralized. The technology to decentralize sequencers is ready, he says, and the people at the top of those companies are not taking the steps to make those changes.

As CEO of StarkWare, which provides technology for Ethereum L2 Starknet, I am one of those people that Resnick is talking about. I can only speak for StarkWare, and not for other layer 2s, but we know it is very much in our best interests to fully decentralize the sequencer. 

Resnick isn’t entirely wrong in his points. StarkWare earns revenue from the sequencer — though this revenue currently more-or-less just covers proving costs. But it isn’t the dark, sinister picture that Resnick paints, in which we are conspiring to steal value from Ethereum or from competitors. 

It’s true that we gain a smaller piece of the pie when we decentralize, compared to a future with a centralized sequencer. But the pie itself is bigger by many orders of magnitude. This is because centralized technology is old hat, and L2s are only interesting if we decentralize and offer something new.   

Decentralization is what consumers want after decades of being at the mercy of technology companies that control, own and manipulate their personal data. We are betting that the increasing desire for decentralization will create a huge market for Web3 products, and for solutions that offer transparency. 

If, in a parallel universe, Starknet were to remain forever centralized then yes, we might quickly earn some revenue in the short-term, but we would always be subpar compared to centralized products such as centralized exchanges. Without decentralization, all we offer is clunky duplications of Web2 products, and consumers will flock back to the originals . 

Plus, continually teasing decentralization then offering something else instead is a horrendous business plan. Nobody wants a technology that is masquerading as decentralized. It might work in the short-term, but we know consumers are smarter than that. They are discerning and will vote with their wallets if we aren’t meeting their needs and their principles. We need to decentralize to succeed. If not, then Starknet itself will fail entirely. 

Even if consumers drop the ball and fall for decentralization theater, industry voices such as Resnick will be there to blow the whistle and hold us to account for our decentralization promises. Resnick is building a picture of where our industry is heading if we don’t get our act together and it’s a rightful warning shot.

What’s holding us back? Despite what Resnick says, decentralization is really, really hard. We have had a team of people toiling away at a roadmap for reaching full decentralization for a long time. We have made significant progress, but this is a marathon, not a sprint.

The most recent win on decentralization for Starknet is staking. It’s another stage in moving towards a decentralized sequencer. A week-and-a-half ago the community vote approved plans that are expected to make Starknet the first L2 with protocol staking. This decision paves the way for enabling validators – after a series of upcoming steps by the network – to sequence and validate Starknet blocks, which is where the true magic of decentralization takes effect. With the recent vote we made a major leap in the ongoing process of decentralization, all while our engineers continue working on the infrastructure to become decentralization-ready at the protocol level.

Staking isn’t perfect. We will be rolling it out in stages. It took Ethereum three years to get this right. It’s also going to take us time. Luckily we can follow in Ethereum’s footsteps, but we also face unique technical challenges, such as figuring out how to navigate decentralization alongside proof generation. 

When Resnick spoke on Bankless he didn’t mention Starknet’s staking plan and I don’t blame him. There are 73 Ethereum L2s listed on L2beat, with another 82 in the pipeline, according to a report from Cointelegraph. We probably have too many L2s, as Resnick rightly points out. Some will choose to be parasitic to Ethereum, but every great advance has the hangers-on who are in it for dubious motives. Sure, they taint the space, but there are enough of us who are serious to mean they don’t discredit it.   

Resnick was right to fire his warning shots at us. L2s need to move quicker to decentralize. Voices such as Resnick and Vitalik Buterin, who will no longer publicly mention layer 2s that aren’t “Stage 1+” in the decentralization scale, are holding our feet to the fire.

So, thank you Max. I hope you’ve left us L2s revved up to do everything we can to show that your words are only a warning — but not our destiny.