juicyfx 「🦑」🦭/acc

the @ikadotxyz validators have completed a major upgrade on mainnet today, including the introduction of 2pc-mpc v4! hooray! 🎉 we call v4 "fast schnorr" because it allows us to achieve thousands of eddsa signatures per second, even with over 100 nodes. this is a major milestone, and a testament to how far ahead of the curve ika is, but i actually want to talk about something else a super cool aspect of this upgrade that's enabled by v4 and that is running now on mainnet is the presignature pool, and this is why i think it's so cool: every mpc protocol has multiple communication rounds, some can be offline (presignature rounds) and some online (signature rounds). the advantage of an offline phase is that it doesn't require all parties to be online and communicate, so can help with optimizing performance presignatures can be message independent, meaning you can compute them without knowing what the message that needs to be signed is, this obviously helps with optimizing performance because you can just accumulate presignatures so they're ready for when there's peak demand with 2pc-mpc v4, the ika network (which is essentially one of the 2 parties in every dwallet together with the respective user) can generate message-independent presignatures. but it goes even further, and generate *client* independent (and as a result public key independent) presignatures, not only for schnorr (live since v3) but also for ecdsa. what are client-independent presignatures? they are presignatures that the network can compute and can be used with any dwallet, for any message. starting today, the ika network computes presignatures in the background continuously, storing them in a pool, so it is never idle. whenever a user requests a signature, they can pay for presignatures from the pool, and essentially only do one online signing round, dramatically reducing the latency to a single round (~400ms), and allowing the network to support peaks of activity without creating bottlenecks. it's hard for people who aren't deep in the weeds of threshold cryptography to appreciate what was required to get here, the level of flexibility and scalability is science fiction when compared to any other threshold signature infrastructure. but the end result is that ika is the only mpc infrastructure in the world that can scale to support high volume use cases (like payments for example) for millions or even billions of users, and the cherry on top is it doesn't even need to sacrifice decentralization to do that! btw, we dubbed 2pc-mpc v5 "fast ecdsa" because it will allow us to achieve thousands of ecdsa sigs/second, which is something even centralized mpc providers can't dream of. it will also be the first single-round mpc protocol in proudction, as the signature will be computed with fhe (using refhe). more on that in the future.

squidfra #3 - accountfi ---------- besides dramatically improving the state of the art in threshold cryptography, @ikadotxyz introduced the novel dwallet primitive, which is the first zero-trust programmable and transferable decentralized signing mechanism. being able to transfer any account on any blockchain is something that was never possible before, and it creates a new type of financial category in crypto - accountfi - which will be possible exclusively on @SuiNetwork if you want a refresher on how account transfer technically works with ika, check out this post: nitter.cf/d3h3d_/status/19303845… ---------- 1. problem it solves accountfi is a new category, so it will take time to explore the new possibilities it opens up, however there are 4 main problems that can be identified now where account transferability is extremely useful. the first is price discovery where a market can't be created: illiquid assets, intrinsic account value and large/complex portfolios. today there's no way to have a marketplace where locked tokens, soulbound tokens (sbts), vetokens etc. are freely traded. same goes for wallets farming airdrops or gaming accounts with their in-game assets and history or entire multi-chain portfolios. being able to trade an account allows the market to efficiently price basically everything. even things like entire daos or rollups. the second is liquidity. while price discovery is important, being able to make everything liquid in a simplified and secure way is also a big problem. the previous examples are pretty clear, but we can also look at something like gaming as an example. there's a grey market today estimated at over a billion $ annually of buying and selling gaming accounts (including "boosting"). this area is rife with fraud as users share their login credentials and pay with gift cards, but when people go to those lengths and take that level of risk to solve a problem, it just highlights what a big problem that is. side note: as the introduction of accountfi solves problems around efficient market pricing and liquidity of assets, it also creates problems for a lot of existing designs because it breaks their premise. sbts were designed to be non-transferable, the sui kiosk was designed to enforce royalties, locked tokens and vetokens were designed to be illiquid, airdrop designs try to reduce speculating farming activities - accountfi creates the need for more advanced designs, that consider the fact the account itself might be tradable, and this need actually creates more opportunities for accountfi squidfra. the third problem is the gas efficiency and finality of legacy networks (like bitcoin for example). accountfi squidfra allows devs to adopt novel designs where they transfer the account ownership (which has sui+ika low cost and sub-second finality) instead of performing expensive actions on legacy networks and have a long wait for finality, which can be catastrophic in some use cases (think instant native bitcoin collateral liquidations instead of waiting 60 minutes). special shoutout here for the 7 day finality problem of optimistic rollups - accountfi squidfra can casually solve one of ethereum's biggest problems. the fourth is provenance - which might possibly be the biggest problem of all. it's practically impossible to keep track of asset ownership in the blockchain, especially across multiple networks and in networks without smart contracts like bitcoin. accountfi allows for provenance of assets ownership through tracking and enforcing policies on the account owners. that is huge for institutional adoption, since aml regulation and sanctions aren't going anywhere - provenance and proving source of funds are critical problems that don't have a good solution in crypto, and that accountfi can very elegantly solve. ---------- 2. use cases it enables accountfi squidfra can power new designs for existing use cases like trading (transfer accounts instead of tokens) or lending (use an account as collateral), but it also enables many novel use cases. the most obvious novel use case is an account marketplace. being able to buy or sell any account including all of its assets, history and activity. it stands to reason that people will model their activity with this marketplace in mind, creating separate dwallets for different actions and purposes, so that they can be easily priced and sold. another novel use case that i believe can be huge is proof-of-ownership. tracking owners, and potentially limiting ownership to parties that have been vetted in some way (like @HumnPassport proof of clean hands or perhaps full-on kyc/kyb) - will make it very easy to have deeper institutional adoption of crypto. today for example institutions pay bitcoin miners a premium for "virgin bitcoin", which is freshly mined utxos that didn't have any other owners. proof-of-ownership can keep bitcoin "virgin" forever, while enforcing a policy and verifying the full chain of custody adhered to that policy. ---------- 3. existing alternatives today, the only way to transfer an account, is by using some account abstraction setup. that is limited to networks that have this kind of setup and of course only applies to a specific account on that network. @NEARProtocol is a good example of a network that allows that out of the box for near accounts, you can trade account abstraction near accounts and there's even a near account marketplace - @NameSkyApp. theoretically, a legacy mpc network could create an account, give user a the permission to use it, and then move the permission to user b - but as you can already understand, this is a far cry from transferring ownership. in fact, neither user a nor user b ever had ownership, all they had was the permission to do things for an account that is owned by the mpc network. in any case most legacy mpc networks don't go that route anyway because of speed and scalability issues, and instead just operate a single account for many users, managing an internal ledger, and only allowing them to do predefined actions that are supported by their internal ledger. ---------- 4. competitive advantage with ika the whole concept of accountfi is revolutionized and enabled by ika - the novel design of the 2pc-mpc protocol that enables programmable and transferable dwallets makes it possible. this means that sui will be the only place where universal and generic accountfi is possible. besides zero-trust security, the superior speed and scalability of sui and ika also give an edge to accountfi squidfra being used in interoperable defi use cases - sub-second finality for native bitcoin transactions for example is a game changer. ---------- although transferring an account is a built-in ika capability, being able to use it for different use cases is not as straightforward. there are complexities around things like the user share management or connecting between the account and any underlying assets and activity it may have across multiple chains. solving these complexities for devs is low hanging fruit for anyone wanting to build squidfra. i'm super excited about dwallet transferability. i think it's one of the things that will open up an entirely new design space and will reshape crypto, and having robust accountfi squidfra is going to be critical for that. ---------- in the next post, we'll talk about ai-focused squidfra - making decentralized access control work in a digital world that is moving towards ai agent domination