This is the fourth of four blog posts on the current state of science and the promise of web3 technologies to improve it.

Part 1, A Credibility Crisis in Science, discusses how the incentives of scientists in the current system contribute to the widespread replication crisis in science.

Part 2, The Business Model of Scientific Journals, explains the problems arising from the current business model of scientific publishers.

Part 3, Curating Scientific Manuscripts for Truth, asks the normative question of what scientific journals or DeSci environments should focus on in the evaluation of research papers to advance science.

Here, in Part 4, we discuss how web3 technologies can offer a path to substantial improvements in science.

Science is the foundation of knowledge and progress (1). As an essential public good, science needs to be reliable, transparent, and openly accessible. The first three parts of this blog series discussed how the current state of science needs to be improved upon on all of these dimensions. Web3 technologies offer exciting possibilities to change the status quo and a new decentralized science (DeSci) movement is beginning to harness these new technologies to improve science (2,3).

Technological innovations have historically enabled vast improvements in our ability to produce, verify, preserve, and disseminate knowledge. Examples range from the invention of the printing press in the 15th century, which first made disseminating knowledge at scale possible, to the invention of the internet in the 20th century, which dramatically reduced transaction costs on sharing information.

The latest wave of innovation along these lines concerns human coordination and value redistribution at scale using web3 technologies, which aim to enable a decentralized version of the Internet. Web3 is based on peer-to-peer networks of a growing list of self-certifying, massively multi-client databases called blockchains.

Just as the early web of the 1990s, Web3 is now in its experimental phase. Skeptics rightfully point out that the space is currently rife with scams and excessive marketing hype. However, there have also been significant technological breakthroughs which could be used to create a better, more open, more verifiable and secure scientific record.

There are at least six ways we have identified in which web3 breakthroughs can benefit science and scientists by enabling:

1. New ways to cooperate
2. More transparent, changeable incentive structures
3. Decentralized data and meta-data storage
4. Verified research objects
5. New forms of identity management
6. New ways of funding research

Below, we consider each in detail

New ways to cooperate

The core promise of web3 is the widespread redistribution of value to users from the efficiency gains stemming from the trustless, censorship-resistant execution of code orchestrated through a peer-to-peer network secured by the incentive alignment of stakeholders. Built on top of these networks are decentralized autonomous organizations (DAOs), i.e. communities of contributors and active users that cooperate through co-ownership and co-governance of these networks.

More transparent, changeable incentive structures

The programmability, composability and trustless execution of smart contracts has created a rich experimental space for mechanism design. Transparent incentive systems can be programmatically orchestrated by a system that does not depend on trust placed on an intermediary. All events are recorded on an open, immutable ledger. These self-certifying, massively multi-client databases can be used to improve accessibility, transparency, and the trustworthiness of science substantially.

Decentralized data and metadata storage

The web3-native interoperable data models such as IPLD, in conjunction with storage incentive layers (e.g. Filecoin) enable making science permanently available to everyone. Persistent identifiers (PID) on web3 are replaced with Content identifiers (CID). Unlike PIDs, CIDs are cryptographically secured to provide a tamper-proof mapping to their underlying content.

Unlike DOIs (the publishing industry’s PID and primary key) which are expensive to maintain and depend on centralized authorities controlled by the publishing industry, cost money, and need to be constantly crawled for repair, web3 CIDs can be issued at will, are not subject to link rot, and are immune to content drift. The transition from web2 PIDs to web3 CIDs would be a paradigm shift in our ability to produce richly interconnected, immutable knowledge graphs (CID-graphs) and automatically generate rich scientific metadata from these relationships. Such as web3 knowledge graphs would reveal new information about connections within the research world (4).

Verified Research Objects

A shift to web3 would also be a step to overcome the outdated legacy of the static print and PDF technology for sharing scientific knowledge towards rich, dynamic research objects (5) that combine text, code, data, videos, peer-reviews, annotations, and commentaries, allowing to trace the development of a research object over time including version management, citations, replication status, and the on-chain verification of open science practices, badges (e.g. for FAIR data), and replication status.

Research objects would be bundled with reproducible code dockers, and allow for the systematic reconstruction of research results via reproducibility runs. Decentralized compute solutions combined with fully homomorphic encryption would allow users to share sensitive data in a privacy-preserving way, thus opening the door for widespread reusability and protecting commercialisable scientific IP.

New forms of identity management

Web3 also empowers pseudonymous identities that can be tied with individual contributions to science across platforms in a tamper-proof and auditable way (e.g. authorship, peer-review, scientific commentary) (6). By combining “proof-of-skill” with pseudonymity, it is possible to create a science ecosystem that simultaneously promotes open debate and reduces bias.

The scientific expertise of pseudonymous contributors can be proven with zk-SNARKs connected to a decentralized identity management system. Such a system would allow credit to backflow to the scientists while providing sufficient bits of pseudonymity to prevent identification.

New ways of funding research

DAOs enable innovations in how crowdfunding is organized. For example, VitaDAO collects funding from its members for longevity-related research that could lead to start-ups or patents. In return for the research funding, the DAO members become (co)owners of the intellectual property that arise from the research. Unlike the traditional financial industry, web3 offers programmable funding channels that can be combined with research outputs and securely preserved in escrow accounts until certain conditions are met.

For instance, funders can announce a research question or program in a node and dedicate money to its wallet. Other parties can easily add funds, pooling contributions from different sponsors efficiently. Research teams apply, their proposals get evaluated, and funding gets matched. Such a system would be faster, more efficient, and more flexible than ordinary grant schemes, allowing a quick response to new challenges (e.g. a new pandemic). Such a model has been tested successfully in the context of Gitcoin grants.

Thus, the DeSci movement can support efforts within the scientific community to change how research is funded and knowledge is verified and ultimately shared, and help shift the ownership of scientific knowledge and its value away from intermediaries back towards the scientists who created that value.3

Table 1 below summarizes the main differences in how the scientific record is managed in the traditional system versus web3 (DeSci).

Table 1 - The scientific recordDeSci Nodes: Reproducible Research Objects + Funding Accounts

By indexing a research object on-chain and linking them to financial channels, we create a new primitive which combines incentive programmability with the production of scientific knowledge as dynamic research objects.

For instance, we can automatically create escrow accounts to channel funds securely between patrons, authors, research institutes, foundations, DAOs, and for-profit R&D at scale. These accounts are open to anyone to create validation grants: peer-review grants, artefact evaluation grants, replication grants, and direct funding for conducting a vetted analysis plan. Unlike the traditional banking system, all financial transactions are recorded on-chain providing total transparency and audibility.

Coupling on-chain research objects, tamper-proof data stored on IPFS, and trustless escrow-based accounts, we can create a 1-click funding solution open to everyone, with complete efficiency, traceability, and accessibility addressing the entire space of scientific validation with surgical precision.

We call this concept “DeSci Nodes” - as they are inherently native to the web3 stack. This has the potential to dramatically increase the speed of scientific validation and speed up funding for research, which would be especially useful to find quick, reliable responses to new challenges, for example in times of emergency.

Selection and validation of science

A core feature of the scientific enterprise is the selection of research articles into journals and conferences based on peer-review. Despite the disadvantages of the current system that we discussed in the first three blog posts, it will remain crucial for scientists to secure funding from institutional sources. The traditional system is here to stay because of the prestige it carries and the sticky incentive systems that universities, research institutes, and funding institutions have created around them.

DeSci can complement and improve on the status quo by offering on-demand validation services of research objects and artefacts (e.g., peer-reviewed ACM badges, OSF badges, FAIR data). Furthermore, replication studies can be explicitly incentivized, and the replicability of results can be independently validated and indexed on-chain. This holistic curation service is performed in the open domain, without claiming copyrights or other intellectual property rights on the research object. This would enable scientists to keep publishing their manuscripts in the existing journals that are most relevant to them while simultaneously making validated research objects open and free for everyone.

Securely stored and verified research artefacts can also help editors of scientific journals to evaluate the submissions they receive quicker and better. On-chain proof of reproducibility, verified badges for FAIR data and open science practices would signal to editors which submissions are most likely to report true, replicable findings. This could assist ambitious submissions to be successful, reduce fraud, and speed up the turn-around time of journals.

Scientific societies as DAOs

One way to orchestrate validation of research objects indexed on-chain (peer-review, artefact evaluation, replication evaluation) is for scientific societies to run a DAO stack (essentially, an OSS preprint validation overlay for coordinating micro-tasking that interfaces with the chain storing the research objects and distributes credit to scientists).

Such DAOs should be run by reputable scientists in each field and enable them to monetize and share the value they create with their services with each other, for example by collecting validation grants, and redistributing them automatically in the form of fellowships with stipends to early career researchers participating in the DAO.

DAOs that return value to scientists

Devising sustainable mechanisms to create value redistribution away from Big Publishers and towards decentralized scientific communities is now urgent. Talented scientists around the world are in need of new ways to uplift their economic conditions. Structural factors such as rising inflation with stagnating wages coupled with increasing consolidation of grant money allocations towards boomer PIs have created an unsustainable financial situation for early-career scientists. The situation is so dramatic that even a few hundred dollars a month can make an extraordinary difference in the life of a talented early-career researcher.

DeSci Labs is building ARCs for that purpose.

Table 2 below summarizes the main differences in how research is evaluated and selected in the traditional system and in DeSci.

Table 2 - Curation of the scientific record

Recoupling impact and replicability

The current scientific record is biased towards novelty at the expense of replicability in many academic fields (see Why Science is Facing a Credibility Crisis). This bias undermines the trustworthiness and legitimacy of science and imposes substantial social costs. One of the core reasons for the replication crisis is the lack of explicit incentives for scientists to share research outputs beyond their manuscripts, carry out replication studies, and think about the replicability of their novel discoveries from the outset. DeSci can change this by creating explicit replication incentives that reward scientists for conducting well-designed replication studies and for publishing results that can be replicated by other teams.

Towards a world where knowledge is openly reproducible, verifiable, and accessible - and returns value to scientists

In summary, the vision we have outlined here would improve the reliability, transparency, and accessibility of research. It also allows scientists to capture a part of the value that they create with their work. Furthermore, it allows institutional and private patrons of science as well as the private sector to direct funds to topics, teams, and approaches that they perceive as most valuable and promising in a fast, capital-efficient, and flexible way, with total transparency.

Authors: Philipp Koellinger (A,B,C), Christian Roessler (D), Christopher Hill (A,B)

A - DeSci Foundation, Geneva, Switzerland
B - DeSci Labs, Wollerau, Switzerland and Amsterdam, The Netherlands
C - Vrije Universiteit Amsterdam, School of Business and Economics, Department of Economics, Amsterdam, The Netherlands
D - Cal State East Bay, Hayward, CA, USA

References

1. Deutsch, D. The Beginning of Infinity: Explanations That Transform the World. Penguin Books (2012).
2. Hamburg, S. Call to join the decentralized science movement. Nature 600, 221 (2021).
3. Hamburg, S. A Guide to DeSci, the Latest Web3 Movement. Future (2022). https://future.a16z.com/what-is-decentralized-science-aka-desci/
4. Cousijn, H. et al. Connected research: The potential of the PID graph. Patterns Vol 2, Issue 1, 100180 (2021).
5. Bechhofer, S., De Roure, D., Gamble, M., Goble, C. & Buchan, I. Research objects: Towards exchange and reuse of digital knowledge. Nature Precedings (2010).
6. Increasing Politicization and Homogeneity in Scientific Funding: An Analysis of NSF Grants, 1990-2020. CSPI Center (2021). https://cspicenter.org/reports/increasing-politicization-and-homogeneity-in-scientific-funding-an-analysis-of-nsf-grants-1990-2020/