Computers do three things, and only three things:
- Store data
- Compute over data
- Send data to other computers (who in turn store data, process data, and send data to other computers, etc.)
You cannot compute without storage, and so naturally most of the earliest Web3 protocols were focused on storage. This includes Arweave (in which we are investors), Filecoin, and Sia.
We have been thinking about the Web3 compute sector for years, and have made a number of investments: Livepeer, The Graph, and Render Network. Notably, all three of these Web3 compute primitives are highly specialized around a very specific type of computation.
Today we are excited to announce our latest investment in the category: Fluence, in which we recently led a $9 million round alongside Alameda Ventures, Tiger Global, Protocol Labs, Arweave Capital, Polymorphic Capital, OP Crypto, Signum Capital and UOB Venture Management.
A simple mental model of Fluence is decentralized and permissionless AWS Lambda. Fluence can read data from any public data source—IPFS, Filecoin, Arweave, Ceramic, Ethereum, Polygon, Solana, Flow, etc.—compute over it, and store the newly computed data back in any of these repositories.
Additionally, as opposed to blockchains where all nodes replicate the same data, Fluence manages the execution over a peer-to-peer (p2p) network with customized decentralization, fault tolerance, and cost on a per-function basis.
Fluence is powered by a custom programming language called Aqua, designed for building distributed systems in a trustless environment and managing the execution across a p2p network. Application developers can use Aqua to create custom p2p algorithms for data replication, computation verification, failover, and load balancing. Aside from scalable computation, this enables the design space for p2p systems that are easier to be built and seamlessly composable with each other.
What is Fluence useful for? Computations that need at least one of the following:
- Need to be verifiable in the public domain
- Cannot have a single point of failure
- Need to be censorship resistant
- Are too heavy to go on a blockchain directly, but where the outputs of the computation must be stored in a public ledger.
Here are some concrete examples of these applications in practice:
- On-Chain Voting — Users sign transactions and send them to the Fluence network, which will count and aggregate all the votes and submit a finalized vote to the chain.
- Mutable NFTs — There is a vast design space for designing NFTs that change over time, or that allow for recombinations and other mathematical relationships.
- Games — As an increasing number of games leverage fungible and non-fungible tokens for in-game mechanics, there is going to be an increasing amount of computation between those pieces of state and gameplay that users see. Today, many of these computations are performed in a centralized manner (e.g., in Axie). As more people come to earn their income from in-game crypto economies, we expect users will increasingly demand transparency on these computations and they will shift from centralized venues to decentralized venues like Fluence.
- Cross-Chain Computations — Automated transfer of an asset from one chain to another requires a worker to detect the event on the initial chain (e.g., token transfer or NFT burn), generate the corresponding data, and send the resulting transaction to another chain to mint an asset there.
- Smart Contract Automation — Fluence can perform the whole set of functions required for DeFi and DAOs, including limit order execution, automated liquidity provision management, protection from debt position liquidations, DAO proposal execution and software updates.
- Oracles — As Fluence provides a full featured framework for creating decentralized systems, it can be used to create sub-networks powered by consensus or any other data verification model. Developers may set up sub-networks to provide data feeds on-chain and apply custom trust models.
- Off-Chain p2p Coordination and Multi-Sig Wallets — Fluence provides a strong foundation for threshold signatures setups and multi-party computation.
Fluence embodies an extremely bold vision for the future: a logically centralized computer that can scale infinitely and coordinate and process any number of functions across any data inputs in a permissionless setting.
This year Fluence will launch an on-chain marketplace for hosting and compute resources and for code. Code authors will share in hosting revenue based on the use of their code modules. The code, when uploaded, is immutable, preventing malware from being inserted later and as long as the modules are being used and hosted, the dependencies will remain functional without the risk of being taken down by a centralized party. As developers upload more code, all of which naturally composes using Aqua, the Fluence network compounds. Composability is the greatest substrate to facilitate compounding.
Moreover, Fluence enables the ultimate dream of open source: get paid for writing high-quality, open-source code that others can use. This monetization model is at the core of the Fluence ethos and will help re-shape the way we think about value capture in open-source systems. Developers have been dreaming about this monetization model since the advent of open source 30 years ago; Fluence is making it possible.
The SDK is available now, and there are already thousands of services running across the network. If you want to help build the substrate to scale Web3 computation, Fluence is hiring, and you can meet up with the team at ETH Denver this week. Reach them at @fluence_project on Telegram and Twitter at @fluence_project.
Disclosures: Multicoin has established, maintains and enforces written policies and procedures reasonably designed to identify and effectively manage conflicts of interest related to its investment activities.