Since Kosmos’s inception in 2017, we have witnessed a Cambrian explosion of heterogeneous programmable settlement layers that offer different balances of decentralisation, security, and scalability as a response to the fundamental problem of execution redundancy in a shared global state.
“ The internet of money should not cost 5 cents a transaction”
Kosmos was founded on the fundamental thesis that the presence of these different properties of consensus, both technical and socially, would enable a concentration of value at the protocol much grander than that of the application layer (all shall hail EIP-1559)
This motivated our seed round investments in alternative blockchain architectures including Solana, Polkadot, Oasis, and Algorand , amongst many other prolific protocols while enabling our founding team to intimately grasp the technical underpinnings of the most fundamental unit in how transactions are processed according to specific state transition functions, mechanism designs (incentivising parties to participate in state maintenance), jump-starting the ‘cold start’ via products/community-based network effects and tokenology.
One thing is certain, network bootstrapping is hard. It requires more than novel ways of reaching consensus on a piece of state or a unique programming language design, but rather a pattern of social consensus and community efforts that create feedback loops that are meaningfully valuable and unattainable in walled garden systems. Users now evolve into owners, creating robust coordination methodologies that incentivise action toward a shared goal. This is the true power of crypto that is grossly under-indexed on the mindshare of our traditional counterparts and their definition of a ‘network’.
Now, we enter the next iteration of Kosmos and transition to ‘Zero Age’, a target raise of a $60m pre-seed/seed fund focused on investing in visionary founders building industry-defining protocols. Our thesis-led approach runs deep and is informed by our in-house engineering and research team which will be routinely sharing insights into what we invest in and why.
“We believe distributed consensus and the abstraction of trust will conjure the largest unlock of economic potential since the advent of the internet”
Source: 2022 Electric Capital Developer Report
Although the industry has experienced many layer ones (L1s) conjuring large pools of collective economic wealth (TVL) and house hundreds of compelling applications, smart contract development is nascent and the developer numbers we witness (w) today are rounding errors in the context of traditional computing systems and software. For one to conclude the further perusal of alternative approaches to the infrastructure ‘stack’ would be wildly uninformed and detached from historical market structures and dynamics. Rome wasn’t built in a day and neither will the desired properties of distributed systems- scalability, liveness, and censorship resistance. It takes time.
And yes, the term ‘infrastructure’ is wildly unbounded and synonymous with many different verticals. We specifically see industry-defining value in the following vectors;
One large decision for any blockchain (and ultimately dApp devs) — what is the format of transactions?
Although the EVM is the most adopted runtime, it is by no means the most technically optimal. In a highly elastic and resource-constrained environment the budget for execution is trivial given the bulk of a ‘block time’ is spent propagating and agreeing on the availability and official order of transactions. One fundamental limiting factor of the EVM (on ETH L1) is its existence as a single-threaded and serialised runtime environment. That is, the state of the blockchain must be paused until the previous transaction is executed.
For the uninitiated, a Parallelised runtime environment enables transactions to be executed concurrently over different cores (much like modern computing) based on dependencies (does the transaction touch the same piece of state?). For instance, Solana has stateless programs where all ‘state’ lives in external accounts, and those accounts are passed into function calls so the dependencies are specified ahead of time. Other champions of this notion are Sui’s Move, Fuel’s UTXO access_list approach, and now Monad’s deferred/ optimistic flavoured path toward reducing the dimension of tradeoffs for devs. We expect this design to become ubiquitous with all user-motivated smart contract platforms and layer two execution environments (not leveraging zks recursive properties) in alignment with traditional predecessors such as under-utilised GPUs and symmetric multiprocessors (SMPs)
The core tenant of modularity in the context of blockchains is the decoupling of at least one component of the stack (execution, settlement, consensus, and data availability).
In a resource-scarce setting, it may not be suitable for monolithic clients to participate in both consensus and validity rules while ensuring that data is meaningfully available. Celestia is a prominent bearer of this flag and utilises error correcting codes (Erasure coding) plus other cool stuff like Namespaced Merkle trees and random sampling-based data availability techniques to abstract block verification to data availability verification. In other words, given a fundamental constraint of data propagation and availability, Celestia enables heightened data throughput of 8mb/s as opposed to Eth L1s 1.4mb/ block reconstruction even after EIP-4844. Moreover, having a pluggable DA layer such as Celestia enables sovereignty and expression of where the canonical (correct rollup state) rollup is defined. Without getting too deep in the weeds, this enables the sovereign rollup to locally (by the local peer-to-peer sub-network) verify the rollup block data was included and ordered by any DA layer via a Merkle proof.
Although Celestia has ‘memefied’ modularity, our definition permits interests in shared sequences, bifurcating liquidity layers, and new software modules that lower the opportunity cost of permissionless innovation.
Zero-knowledge can be defined as computational integrity when no one is looking. Put more formally, a validity proof such as a Succinct non-interactive Argument of Knowledge (SNARK) is a type of proving scheme where a party (the prover) proves to another party (the verifier) that a certain computation has been performed correctly. Notice the absence of ‘zk’, which has largely become a misnomer. For a system to be of ‘zero-knowledge’ capacity there must be no information leaked about the witness (w)to the verifier. The true magic of this moon math is that the succinct proof can be verified faster than re-executing the transactions (full circle to the first sentence)
As discussion regarding the thesis on executing native EVM bytecode (or similar) within the EVM (zkEVM) has been conducted with brevity and followed by an ambitious wave of capital. as a fund, we are more provoked by the ability to natively bridge between “fractured trust networks”.
Now, the industry can run, yes run, away from Multisig-type architectures and toward a trust-optimised (too scared to say trust minimised in this environment) communication via running light clients on the source/destination chains and verifying the consensus (keeping track of blockheaders + BLS signatures in the context of Eth) and thus attain confidence in one another’s state. Fluent and trustless omni-chain product experiences will concentrate liquidity and de-fragment a historically jarring user experience. This by no means signifies that we undermine the elegance and significance of the compression and recursive properties of validity proofs for general-purpose computation (LLVM-based approaches or other fully functional languages ) but is an indication of the wider breadth of validity proofs.
A protocol is only as decentralised as its weakest link. Decentralised compute is concerned with distributing computation across many nodes to achieve some increased degree of liveness guarantees (fault tolerance) and open participation. This is a highly overlooked vector and equally important component of the stack for creating composable, immutable, and persistent databases that together are much more valuable than the sum of individual parts. While Filecoin continues to eat away at traditional cloud storage for persistent immutable data, we expect other solutions that leverage zero-knowledge proofs and cryptography rather than centralized access controls, to dramatically improve data integrity, availability, control and the true unlock of composability. That is, the de-fragmenting of siloed data streams to enable collaboration and iterative definition of read and write authorities, increasing the richness and portability of previously impeded information.
The rate of permissionless innovation on the dApp layer is directly correlated to the bootstrapping and inclusion of more vast frameworks and tools that reduce the opportunity costs of building on the world’s new supercomputer… if we truly want 1 billion users on-chain, we must first have 1m developers.
Applications that leverage the unique substrate of blockchains including censorship resistance, credible neutrality, and open access can create collective and programmatic summoning of net-new consumer behaviours. As we move toward a reality of fungible blockspace, protocol-specific decision-making can be abstracted — dApp devs will soon be able to focus on scripting the core business logic of their application rather than making protocol-based tradeoffs, unleashing permissionless innovation and the push to re-define a more open internet at the application layer.
A more open, inclusive, and composable on-chain financial layer that aids in wealth distribution. Here we place prominence on departing from the rental of liquidity and locusts of finite activity toward deep long term alignment between those providing ‘value’ to the protocol and natural holders of the native token. Zero Age expects to promote on-chain financial applications that help abstract away the complexities of opaque markets such as derivatives (‘real world asset synthetics, options and more nuanced instruments) empowering less financially fluent economies ownership over their finances.
Protocols that use sophisticated token distribution schedules and incentives as a medium to channel behaviour toward a common goal/outcome. Ultimately, the token design is a non-linear process that may be iterated but should remain grounded in tethering the usage of the token with benefits to the entire protocol (not localised) and linearly rewarding those that provide defined ‘value’. We nerd out in this arena and cannot wait to work alongside our founders in building out their token economy (rich design space) and attach utility so that users are incentivised to hold, reducing the velocity (MV=PQ) of the token and aiding in long-term sustainability. Proof of Useful Physical Work or the ‘X’ to own phenomena (Gaming elements) that prioritize the product rather than short-termed financial engineering of the token and revenue sharing,when profitability allows, is a dramatically underserved arena for PMF in crypto. In many cases users become psychologically and economically attached, capturing the full dimensionality of what it means for network “stickiness”.
Democratizing and illuminating the effects of Monopolistic Extractable Value (MEV). There are many different approaches to minimizing negative externalities related to MEV. One is programmatically enshrining an end expression of what a user actually wants to achieve. Or put more succinctly an ‘intent’. As a user’s expression to transact is the first initiation of MEV in the supply chain its incorporation at the app layer will have potent effects on how a user interacts with an application.
Tokens form the elixir of ownership in Web3 and are our primary focus as a fund when deploying capital in early-stage projects for numerous reasons. To have the most pronounced outcome on the protocols we invest in, Zero Age will be primarily concentrated on pre-seed and seed projects with an obsession in helping them reach escape velocity via our diverse team backgrounds that hail from engineering, entrepreneurship (building) , and traditional finance.
If you are an ambitious founder with similar grand worldviews please do not hesitate to reach out.