Science is in crisis. More than 70% of scientists have tried and failed to reproduce peer research.Less than half could even replicate their own results.
@DeSciLabs is looking to bring scientific research and publishing into the web3 world with DeSci Nodes.🧵 👇
1/ For decades, the PDF format has been the unit of scientific record—the standard by which scientists share, read, and publish research.Unfortunately, many issues in science can be traced back to the PDF and closed journal system of publishing.
2/ The current system suffers from fragmentation, exploitative publishing practices, link rot, censorship, and more.
3/ PDFs work as a "frontend" for science, showing printed results, but many scientists are interested in the backend.Datasets, code, and computations are scattered across the web, if available at all.A PDF limits storage of crucial info required to reproduce research.
4/ The DOI system that organizes PDFs is obsolete.It optimizes for the needs of publishers, rather than researchers.There is frequent link rot that breaks the system of citations that scientists rely on to discover and cite relevant work.
5/ The science publishing industry is exploitative to scientists.It is a pay-to-read and pay-to-write model.Journals have gone from being quality checkers to determinants of career success.Publishers own the distribution and are gatekeepers to scientific knowledge.
6/ Closed journals have run their course. The badge of "peer-reviewed" is not a strong indicator of research quality or scientific truth.
Countless studies and scandals have shown this:
7/ Surgisphere and hydrochlroquine is another recent example:
8/ Journals are also blinded by status. Researchers demonstrated this by submitting 12 already published papers written by prestigious institutions, but changed the names.
The result?Over 90% were rejected, many for "serious methodological flaws"
9/ Clearly, journals no longer provide enough value to researchers and scientific progress to be worth the high costs. What if we could reimagine how knowledge is shared?DeSci Nodes aim to be the web3 unit of knowledge, expanding what is possible in scientific publishing.
10/ So what is a DeSci Node? Developed by @DeSciLabs and supported by @DesciFoundation, it is an open, composable, on-chain record of knowledge.
11/ DeSci nodes are data containers built on IPLD, secured by CID-graphs, and stored on distributed P2P networks that allow scientists to integrate and compose research assets with their papers.
12/ That's a lot to take in. Let's break down the benefits one by one. In contrast to DOI, which stores information based on location, DeSci nodes are stored by content.
13/ Content is cryptographically secured in a tamper-proof method that also addresses the issue of link rot and content drift.
Referencing research will take you directly to the original content, as specified by the CID with no risk of broken links or edited data.
14/ DeSci nodes let scientists add data and functions to otherwise static PDFs.Nodes can include datasets, code, and other critical artifacts needed to replicate and verify results.
And it's all one unified experience—no need to crawl the web or ask publishers for the info.
15/ This allows for composable science—a gamechanger for data availability and publishing.With nodes, citations can be function calls, making it easy to use elements of previous research—whether it be code, computations, or raw data—in an immutable and verifiable way.
16/ Scientists can run the exact code with the same datasets as their peers to replicate findings.
In an era where most researchers haven't even attempted to publish replication research, this could go a long way in addressing the replication crisis.
1,500 scientists lift the lid on reproducibility - NatureSurvey sheds light on the ‘crisis’ rocking research.https://www.nature.com/articles/533452a
17/ DeSci nodes also expand what we think is possible with credentials. In place of peer review, studies could receive badges for providing artifacts.
Papers can earn reproducibility status when other papers cite and re-use their artifacts.
18/ Citations as on-chain functions would enable a record of contribution and impact. zk-SNARKs could be used to verify the expertise of pseudonymous researchers.
There are endless potentials for decentralized identity systems to be built on top of DeSci nodes.
19/ There are also new ways to fund research.
Nodes contain funding wallets to incentive work on replication or specific research questions.
Funders and stakeholders can contribute to the wallet and scientists can claim the reward for their work.
20/ The future for science is bright. DeSci nodes is working towards a world where research is open, transparent, reproducible, and creates value for scientists.
The permissionless and composable nature of the web3 leave endless possibilities for innovation.
21/ When we mix technological threads, we end up with solutions no one could have imagined initially.
In summary:
- Science has relied on the PDF as the unit of knowledge
- Problems are arising such as poor reproducibility and extractive publishing practices
- DeSci Nodes are a web3 unit of knowledge that make it easy to access, share, verify, and fund research
Thanks to @cairoeth, @_eshabora, and @_nishil_ for feedback on the thread.
To learn more about DeSci Nodes and science in web3, check out:
