'Keanu' Explained: What It Means to Merge Two Ethereum Projects

In the traditional business world, corporate mergers often create behemoths that come to dominate markets and, on occasion, drive political leaders to break them up again. 

This is the story of a merger in the making, but in the decentralized world of blockchain-based projects it might help to set aside all prior assumptions about what a "merger" means. 

Two encryption projects, Keep and NuCypher, both running on the Ethereum blockchain, have begun discussing what they are calling a "hard merge," codename: Keanu. Crucially, this would be a merger of their protocols' functions and communities, not of their companies. 

"It really will meaningfully make what I think is already a decentralized network already much more decentralized," said NuCypher founder MacLane Wilkison.

It's not as if blockchain mergers haven't happened. Andre Cronje's robo advisor for yield that is the behemoth Yearn Finance announces a "merger" or a "partnership" every month or so, but it's never quite clear what's going on behind the scenes.

"Instead of Andre announcing that Sushi is part of Yearn and none of us knowing what that means, maybe people could start showing us the code and combining projects," said Matt Luongo, founder of Thesis, the company shepherding the Keep project.

In blockchain projects, companies build protocols but lots of disparate entities all over the world do the work (the kinds of work being limitless) of that protocol and profit from it. The theory is, the more entities doing it, the safer everyone using it is from theft or manipulation. This is called decentralization. 

So the idea in this hard merge is that eventually, all the operators doing work on Keep and the ones doing work on NuCypher might do the work of both networks. It's like a hardware shop with a locksmith inside: certain businesses go well together.

As of now, the Keep protocol has been entrusted with $243 million worth of crypto assets to run its Bitcoin network, according to data site DeFi Pulse. NuCypher has roughly $265 million worth of ETH staked to run its encryption network.

Back in the heady days of initial coin offerings (ICOs), CoinDesk compiled a list of seven projects that investors liked ahead of their sale. Keep and NuCypher made the cut.

Keep described itself at the time as a privacy layer for Ethereum. NuCypher provided an encryption delegation service.

Here in 2021, we have seen and continue to see the higher-quality, better-vetted ICOs bring products to market, and that includes Keep and NuCypher. Both of them have built what are called "threshold cryptosystems." These systems break up cryptographic keys across multiple nodes on a network such that it takes some threshold number of nodes to sign for an action to be validated, spent or decrypted (as the case may be). 

This has twin benefits. First, trustlessness: No one node can take unilateral action. Second, resiliency: Let's say it takes six nodes to collaborate for a valid signature. If that were any six of, say, 36 nodes, then a signature could still be made valid even if two-thirds of the nodes suddenly went offline for some reason.

The combined project will be looking for much larger numbers than that.

"Threshold cryptography in general fits very well into this context of blockchains which is all about trust-minimization and censorship-resistance," Wilkison said. "Basically what we in our communities view is probably fairly likely to happen is: As each network starts to support more and more threshold cryptography services, they will probably look more and more alike."

Neither team brought its product to market until last year. The Keep project was spun out to a new company called Thesis, and opened with a more decentralized way to lock up bitcoin and use it for decentralized finance on Ethereum. NuCypher went live in October following its "WorkLock" distribution event.

Founders at both teams hope they have innovated a scheme for what it actually means to merge blockchain projects.

"It is, as far as we know, pretty novel," Wilkison said.

Meanwhile, the companies behind these protocols would keep developing software to do their respective work, but on Keanu. It's like how email is one protocol but there are lots of different email providers and lots of different software for reading and sending email.

Yes, two networks might become one so that sounds like consolidation. But it might not be. Standard Oil or the WWE might not be the right mental model to apply to blockchains.

Both companies famously hoovered up smaller competitors providing the same service until they were effectively monopolies. "I thought we would see more of centralized companies eating up decentralized companies before this would happen, like FTX acquiring Ren" last month, said Paul Veradittakit of Pantera Capital, an observer of token projects.

But is it the same thing if blockchain protocols, not the companies, start to merge?

"I don't know if you get meaningful decentralization out of many different networks," Wilkison argued.

Whereas the WWE acquired regional wrestling associations and combined them into one national force, blockchain protocols are run more from the ground up by independent operators. It's not the protocols merging that raises centralization red flags.

If the nodes running such protocols were to start quietly combining behind the scenes, though, that becomes a much different story. It's just a story that's much harder to read.

"I don't know that you can look at it as: 'We need many networks.' We need a smaller number of networks and a better distribution of development teams and node operations," Wilkison said.

Brayton Williams of Boost VC, which backed Keep, put it another way.

"In most cases the competition is not crypto to crypto teams,” Williams said. “We are fighting against traditional finance and how things work before crypto emerged. More high-quality teams that can join together to compete against the real competition will make for a better fight."

There had been talk in the NuCypher world of doing a decentralized bitcoin-on-Ethereum solution like the one that Keep has built – in other words, basically competing directly with Keep. 

Wilkinson said the two camps looked down the road and saw years of one-upmanship ahead as each service expanded to encroach more and more on the other.

Both networks use staking to ensure the security of their respective applications. In other words, participants can earn fees for doing computational work, but only if they put some asset at risk in case they screw up or misbehave.

Community members in the space that have gotten into the staking game, however, had begun to expand out to more and more projects. Many of them staked on both networks, and eventually some started to see this coming clash between the two efforts.

As the founders tell it, these community members reached out to the companies to ask if it would make sense to combine efforts.

"No one quite knows what to make of this. and I get it. It's experimental," Luongo said. "We know that at the end of the day we can do more together than separate."

Veradittakit of Pantera Capital agreed, saying, "I think we could see more of these, and it’ll be interesting to see how the models evolve!"

Who doesn't love Keanu Reeves?

So, first of all, it's only a codename. If Keanu becomes real, it will be a decentralized autonomous organization (a DAO), and that DAO can change the name. For now, "Keanu" is a placeholder.

But it has a logic: The movie star's name works as a sort of combination of the two projects’ names (Keep + NuCyphter > Kee + Nu > Kee and Nu > Keanu). They could do it the other way but that would be NUKE, and neither founder much loved that.

Keanu is also somewhat memetically relevant because ... well, okay: another blockchain, Cosmos, did something similar and it was called a "hard spoon." It was a spoon, as opposed to a fork, because it was a big change but also non-contentious. (In open-source software, forking means copying source code and then developing it independently; in crypto, forking a chain is a sort of bloodless secession of one community from another.) With me so far?

The spoon was one network, not two. Also, Cosmos was replicating network state (all the information stored on the system) and that's also not really true here, so ... there is no spoon.

One new protocol will be able to do the work of both networks, and stakers on Keep and NuCypher can opt to join in and do both jobs for more potential fees. It's "hard" because it's a real, definite change, but it's not really a fork because Keanu won't be mutually exclusive with the prior networks either.

Another way to think about it: Imagine this is the early days of the web and one open-source team had created a protocol for sending files from one machine to another. Another open source team had created a protocol for sending text messages from one machine to another.

A hard merge would be as if someone came in and took both codebases, slammed them together and then created email: text messages that can do attachments.

A hard merge is sort of like that, but in a blockchain context such that all the computers in between get paid with a token.

If Keanu launches, then either Keep’s or NuCypher’s token can be used to stake as a work token (kind of like a taxi medallion, in that you pay an upfront cost for the right to provide services) on the new network. In the current proposal, from the perspective of Keanu, one KEEP will be the same as one NU, and vice versa (and for crypto tokens, they are pretty comparable networks in terms of value anyway).

"I think that kind of fits more with the ethos of both communities and the space as a whole," Wilkison said. "If for some reason this hard merge did not work, it's possible for one or both networks to say this isn't working out and we're just going back to our own network."

If it does work, though, the Keanu DAO could also make a move to increase stickiness. For example, it could create a token that's emitted for stakers on Keanu. For now, though, the only tokens are KEEP and NU.

In theory, the whole will be greater than the sum of its parts.

NuCypher, for its part, won't need to invent a way to custody bitcoin in a decentralized fashion because the Keep team has built that, and its nodes can now share in doing that work.

Keep has a plan to dramatically scale up its approach to custodying bitcoin on Ethereum, with version 2.0, but it needs a lot more nodes than the 200 or so it has now. NuCypher has about 2,000.

NuCypher is still developing the decentralized application (dapp) marketplace that it hopes will prove the killer use case for its fundamental offering. Meanwhile Keep has a dapp (tBTC) that has proven market demand, but its first very safe, very careful version is too capital-inefficient to scale.

It needs many more nodes to run version 2, which will require a lot more signatures. That said, it swaps out the collateral requirement for node operators with an insurance-based approach.

"We're going directly after wBTC," Luongo said, referring to the leading solution for bitcoin on Ethereum, provided by BitGo, which currently custodies $7.5 billion in BTC, as of this writing. "Every bitcoin deposited will be backed by on the order of hundreds of signers."

The plan for now is that tBTC version 2.0 will be the first dapp on the Keanu protocol, with more to come soon.

If merging here works, the teams could broaden their horizons to other blockchains. Luongo said, "We are pretty far along with integrating with Celo," a financial blockchain that runs on Tendermint, a consensus system that's compatible with the Cosmos network.

Cosmos is on the cusp of enabling interoperability between a bunch of blockchains. "I'd love to see tBTC v2 launch on Cosmos. I think that would be really powerful," Luongo said.

First, the two teams need to write a specification that covers Keep's signing and NuCypher's encrypting.

Specifications are human-readable descriptions of a service such that different pieces of software (clients) can operate it. Ethereum's ERC-20 tokens, the building blocks of the 2017 ICO boom, are written to a specification, as are those non-fungible tokens that are all the rage right now. The specification delineates all the things conforming software needs to do, but different clients can be designed to prioritize different functions (much like using Outlook, Apple Mail or Mozilla's Thunderbird, for email).

Once that's finalized, each company will get to work building clients that meet the specification, kind of like how Ethereum has had both Geth and Parity.

"Really, what that means is if one of the dev teams gets hit by a bus the project can continue," Luongo said.

NuCypher has a DAO. It may need to approve a smart contract change in order to make it feasible for node operators to move. That's still being determined.

If it doesn't, the NuCypher devs can just get to work on its clients for Keanu, because the DAO runs the NuCypher protocol but not the company.

Keep has a community-run multisignature Gnosis contract (kind of like the crypto equivalent of a board of directors).  Multisig members will  put together some kind of process, which may require node operators to signal support and maybe for KEEP holders to vote.

Both founders want it to move fast if it moves, aiming for a second-quarter release.

"Both sides have to decide: Are we stronger together?" Luongo said.