🏰 Keep: The Tech Bringing Bitcoin To Ethereum — Issue No. 89

This week, the Thesis team, which is behind the TBTC project, announced they had raised $7.7 M via a token sale. TBTC is a project which aims to create a trust minimized bridge allowing Bitcoin to be transferred to the Ethereum network. The project was announced in August 2019 and has since shipped to testnet. The mainnet launch is expected sometime in the next few months. Link.

One question immediately comes to mind: what token did Thesis sell and why is it needed? It turns out, TBTC will be the first project to leverage a new technology called "Keep," which is also built by Thesis. The Keep system includes a native token.

In this edition of Build Blockchain, we'll quickly review TBTC and learn a bit about the Keep project which makes it possible. We'll also discuss what both projects might mean for the crypto landscape moving forward.

Bitcoin On Ethereum

If successful, the TBTC project will make Bitcoin available as a standard ERC-20 token on Ethereum. This would enable Bitcoin holders to participate in the emerging Decentralized Finance ecosystem. The tBTC token (yes, the token is referred to with a lowercase "t") could be added as a form of collateral in MakerDAO, for example. This would allow Bitcoin holders to take out loans denominated in Dai.

There are already forms of tokenized BTC on the Ethereum network, most notably WBTC. What makes tBTC different is the degree of third party trust involved. WBTC is secured by a trusted federation of signers, whereas TBTC aims to be significantly more trust minimized. I covered this in issue number 60 of this newsletter, where I wrote the following:

How does TBTC achieve a trust-minimized bridge? Through a system its creators refer to as a "bonded, multi-federated peg." In this system, anyone can become a "signer" that participates in securing deposits on the Bitcoin network and minting tBTC on Ethereum. Each deposit requires a group of signers to participate, who are chosen randomly. The key is that signers are required to collateralize each deposit with ETH on the Ethereum network, and the bonded collateral can be seized by the smart contract system in the case of bad behavior. Link.

That paragraph describes one of the most important aspects of the TBTC system: using ETH held in a smart contract as collateral for the locked Bitcoin. Astute readers, though, will notice I glossed over some really important details. How can signer groups be formed in a random, trust minimized way, and how do they coordinate without the risk of collusion? These are extremely hard problems. That's where Keep comes in.

Keep It Secret

The Keep system allows smart contract users to store private keys off chain with a group of third parties, or "providers." Users can then request providers publish messages signed with the secret keys on chain for use within smart contracts. To achieve this in a trust minimized way, Keep leverages both advanced cryptographic techniques and token based incentive mechanisms.

On the cryptography side, Keep leverages a secure multi-party computation scheme known as Shamir's Secret Sharing. In short, providers form groups where each store a share of the private key, but none of them have the entire thing. Revealing any one share, or even a number of them, does not jeopardize the security of the key. Though no single provider has the full key, they can still collaborate to publish signed messages back to the smart contract as long as some threshold of them cooperate. Link.

While this cryptographic scheme makes it possible for providers to work together, it doesn't give them any incentive to do so, nor does it dissuade them from colluding maliciously. To deal with these issues, Keep introduces a native token. Contracts leveraging Keep pay their providers for the service via this token, creating an incentive. To participate as a provider, one must also stake a certain number of these tokens, making sybil attacks expensive. Presumably, these two uses will give the token value proportional to demand for the network's services.

Finally, to reduce the likelihood of malicious collusion, Keep needs a way for random staking providers to form groups that collaborate to secure a given private key. To do this, Keep uses a decentralized randomness beacon called a "Threshold Relay" to securely publish a random number in each Ethereum block. That randomness is used to assign providers to groups, preventing entities from intentionally forming a group together and colluding. This technique was pioneered by the Dfinity blockchain, and is used in that network's consensus algorithm. Link.

Trusting In Trust Minimization

By combining advanced cryptography, tokenized incentive design, and decentralized randomness generation, Keep hopes to provide the ecosystem with a powerful new primitive. In TBTC, secure off chain private keys will be used to sign Bitcoin transactions which lock the BTC that is then tokenized on Ethereum. This is a great first use case, but there are many other ways that Keep as a protocol can be leveraged if it proves itself.

Here's one example. Since Keep can store private keys that can be used for encrypting and decrypting files, it should be possible to design a marketplace for digital goods where payments made on chain trigger decryption of files stored off chain. This would function something like a fully decentralized DRM service, allowing content creators to sell video or audio files automatically and without counter-party risk on either side of the transaction. Many more useful applications can be imagined along these lines.

For the moment, though, Keep still needs to make its debut as the technology underpinning TBTC. This is no small use case! Long time readers of this newsletter know that I'm a fan of both Bitcoin and Ethereum, that I believe both networks make logical tradeoffs to achieve different ends, and that a trust-minimized bridge between them will be a boon to both networks. I'm really excited to see TBTC come to mainnet sometime soon.

With all that said, I'm keeping my expectations pretty low when it comes to adoption of TBTC in the short to medium term. As mentioned earlier, we already have a custodial version of Bitcoin on Ethereum in WBTC. It's been around for a while, but adoption has been relatively low. At the time of this writing, there are less than 1000 WBTC in circulation.



Arguably, the reason WBTC's adoption has been limited is because of the associated trust assumption. Since TBTC greatly minimizes trust in any third party, shouldn't we expect greater adoption?

In the long term, yes. In the short term, though, the very thing which makes TBTC attractive will also defer its adoption. To remove those trust assumptions, Thesis had to build out quite a bit of rather complicated technology. All indications are that they've done a great job, but new tech always comes with risks. My guess is that BTC holders will be hesitant to put their funds at risk until the system has proven itself. Ironically, it will take time before they come to trust this new trust-minimized system.