Decentralized Science (DeSci) is what we call digital communities of global citizen scientists working together to solve challenging scientific problems.
We are actively building the software stack for empowering citizen science communities. Our core team are neuroscientists by training but our stack is intended to be plug-and-play for any digital community requiring evidence-based decision making using crowd-sourced data, cloud services, and decentralized community management.
Many scientific questions are becoming big-data problems that require the storage, sharing, and modification of massive datasets. This is already a complicated problem, but add the plaid-work of data operability standards on various cloud platforms, the cacaphony of global regulatory guidance on personal data, and you'll understand why institutions heft out large sums for their compliant cloud-services.
IPFS stands to solve these problems by offering a public commons for content-addressable content stored on a peer-to-peer network of nodes that host data. The IPFS Distributed Hash Table (DHT) is a permanent pointer to the content stored on the network, allowing important scientific knolwedge to exist on a permaweb of knowledge.
The most significant benefit IPFS holds for DeSci is not the most obvious, but it is likely the most important - IPFS establishes self-sovereignty of data ownership when combined with encryption and linked-lists of access permissions. Ironically, the more nodes storing copies of encrypted data, the more secure and resilient the network of data becomes.
A DeSci Community can use IPFS to:
- Store large datasets and command on-demand wide bandwidth for upload/download 💻
- Create private IPFS networks distributed around the world to permanently store important scientific data 🧪
- Fight censorship and promote open data 🦾
- Store cute kitty pics 🥰
- And as a building block for building decentralized applications (dApps) with data self-sovereignty of the user baked-in ✊🏽
You can learn more about IPFS here. See all the cool ipfs dapps at awesome.ipfs.
The access of any given dataset requires a permission management system that respects personal data regulations, intellectual property, and copyright. The solution to this is a web standard for identity that goes beyond app-specific user tables in centralized systems such as Google, Facebook Login, SSO, or other similar implementations that store identity metadata on commercial webservers.
A decentralized identifier standard (DID) is an index of user-data, typically associated with a unique key, that can be shared across applications on the web. A DID grants the user sovereignty over their data and the ability to set permissions on demand for opting in/out of any service.
The end result is better for both data providers and consumers! Applications can easily integrate user metadata across the web and users can decide how and if they share their data with services.
You can learn more about the IDX DID standard here. The EU Commission has a comprehensive review of decentralized identities here. Check out an implementation of DID with 3box profiles.
Our first instinct as data analysts is that we need hands on data in order to put it through the paces of quality control, preprocessing, confound modeling, maybe synthesis with other datasets.. and that's all before we can deploy it to answer our scientific question.
A much better approach that respects informational self-determination and also lets institutions mantain access control to sensitive data is to bring the computation to the data.
What does this mean? Well, implementations of confidential computing allow encrypted cloud sessions to grab secure personal data from IPFS, run an analysis, and make the final results available to the requestor without ever revealing the data. This way individuals keep their data secure but others can leverage it to perform useful analyses!
You can find more information about confidential cloud computing implemented through iExec and Ocean Protocol.
The output of scientific research generally feeds into the economy and is used to build roads, medicine, and technology to make our society healthier, more efficient, and sustainable. However the profits from commericial application of scientific research does not always flow back to the participants that made the research possible!
In the case of human-subjects research, participants provide data, perhaps receive payment once, but immediately lose control/ownership of their personal data once it is published. A similar case exists for scientists or early-career trainees that spend time curating, assembling, and publishing a dataset for a given wage (in some cases for free) but they do not receive any share of the profits when that data is used by commercial entities.
A DID-managed permissions system for data can let users decide whether they want to freely share their data, and in what conditions they expect renumeration!
Data marketplaces may be used to pair liquidity with tokenized representations of datasets living on IPFS and allow fair price discovery of a data asset to take place. Royalties and permissions are programmable by the user and allow for unique game theoretic scenarios to evolve.
You can learn more about marketplaces for datasets on the Ocean protocol! The Ocean Academy is also a great place to begin learning about this rich ecosystem.
A decentralized community thrives on transparency! Especially when it must self-govern, allocate funds to grow the community, and choose a fork on the road to achieving their goals. This is especially important for scientific communities!
A major bummer (and impediment to innovation!) about legacy academic research is the lack of transparency in grant reviews, peer review of papers, how hypotheses are actually formed and tested.. pretty much the entire scientific process.
A decentralized science community can codify its ethos in public smart contracts that automate things like disbursal of grants, assignment of peer review, or even the various checkpoints from hypothesis generation, to model prediction, data collection, and validation!
Open source toolkits like Aragon allow the rapid scaffolding and deployment of autonomous communities with verifiable voting, transparent treasury management, and the power of automation to publish, query, and monitor scholarly products such as datasets, papers, software, and who knows what else?!
You can find an excellent primer on decentralized autonomous organizations (DAOs) here. Find a list of DAOs on DeepDao.
The DeSci stack is far from complete and does not provide a true statement of value for the roiling potential of decentralization + automation + digital communities coming together to engineer our way around non-science problems keeping us from revving the engine of innovation into its true performance spec.
Check out our issues page to get an idea of ongoing projects if you'd like to contribute!
Have an idea? Want to contribute but don't know how? Chat with us on Discord!