Cooperative governance of (neuro)scientific communities with decentralized open infrastructure

Manisha Sinha, Jonny Saunders, Lina Lina

Funding Area: Community governance / Gobernanza comunitaria

Problem Statement: In the recent past, scientific communities have come to rely on corporate social media and forums for public outreach and scientific communication. This has been beneficial especially for those who cannot afford to publish with open access in prohibitively expensive high impact journals. However, due to the nature of corporate social platforms they tend to be closed-sourced, centralized, and involved in surveillance capitalism. Recent executive decisions made at such platforms have led to poor moderation and decreased accessibility both from product development and community points of view. Facing harassment, noticing a rise in anti-science misinformation, and generally concerned about the ethics of such executive decisions, several communities have sought alternatives. One such viable alternative is the fediverse, a group of federated social networking services such as Mastodon, where several communities have migrated. With the needs of the Neuromatch and neuroscience community in mind, we launched neuromatch.social which is a cooperatively-governed Mastodon instance. Our goal is to experiment with open and decentralized infrastructure for scientific communication. Our vision is to democratize science by lowering the barrier to entry to scientific communities and we believe in taking back ownership of our communities from for-profit corporations. Else we risk constant platform migrations and loss of communities that have been built with a voluntary labor of love.

Proposed Activities: While our proposed cooperative governance structure has been adopted by the community (https://wiki.neuromatch.io/Governance) through a consensus-based decision making process, we are still in our bootstrapping phase. During this phase, we have the following activities planned: (a)Formalization of our bylaws, rules, and norms of the community - within the team and community we have the experience of how physical and digital cooperative spaces work. Some of us have years of community moderation experience from Neuromatch Academy and Reddit. Our bottleneck has been the time required to collectively discuss this with the neuromatch.social community and with our friends at other digital cooperatives like social.coop. With funding, we’ll be able to provide honoraria for the Social Working Group members and invite experts involved in this work. We could aim to achieve this goal within the next 12 months. (b) Improve our onboarding process (https://wiki.neuromatch.io/Mastodon/Onboarding): this activity has already been initiated by the Social Working Group in response to a recent survey where some community members expressed difficulties getting familiarized with various Mastodon features. Current progress is available here https://wiki.neuromatch.io/Using_Mastodon and in some of the pinned posts here https://neuromatch.social/@socialwg. Our team has expertise in survey research and this activity will continue to evolve over the entire course of our work based on community feedback about new features. (c) Launch and run Forum mode with searchable tags for scientific discussions - this work is also an outcome of the internal survey. We’ve found a few open-source options to explore, namely Lemmy, Kbin, and Discourse. Within the Technical Working Group we have the system administration and operations expertise to launch the Forum mode. However, we wish to go a step further by indexing already published work keeping in mind the request from the community to make scientific discussions easily searchable within the fediverse. Our implementation could be as simple as revising the hashtag regex to allow for DOIs to be used as hashtags so that people could find references to the research. One step up would be to resolve all urls that have an associated DOI and make canonical pages/posts for every paper that's mentioned. This could be automated by developing a bot that makes a post for each DOI and adds a new reply to the post whenever it is mentioned. Or it could be something that the Social and Technical Working Groups collaborate on. The technical contribution would be to create an additional "papers" feed for the community to browse through discussed papers. We have the scientific expertise within the Social Working Group and the community for the curation of this feed and the technical experience with Mastodon’s code base such that with appropriate funding we can achieve this within the span of 2 years involving continuous integration of community feedback.

Openness: We welcome all members who share our vision. We plan on improving our onboarding process since we recognize this is a barrier to participation in our community. We extend our cooperative spirit by recommending our sibling community at synapse.cafe to others. We have also had meetings with different communities who could adopt our governance processes and source code to make them meet the needs of their community. We are running a public fork (https://github.com/NeuromatchAcademy/mastodon) of glitch-soc (a fork of mastodon). This open source software runs on the open and decentralized ActivityPub protocol. We use loomio, an open-source software, for consensus-based decision-making. Our processes are transparent as we discuss them with the community on our backend discord (which is unfortunately closed-source and we plan to move away from this after our bootstrapping phase is over). We also announce any updates via the admin and socialwg accounts. Most of our work is documented on our public wiki (https://wiki.neuromatch.io/) and with funding we will be able to devote more time for documentation. We plan to expand our wiki to include a member profile directory for building a contribution graph. Typically scholarly communities rely on publications as a key metric to assess contributions to the scientific community. We strongly believe that all aspects of scientific labor deserve recognition. Our aim with this directory is to give due credit for the work done by members.

Challenges: Most of the members of the social and technical working group have full-time jobs. This means that while we will make progress towards our shared goals, the progress will naturally be slow unless we can raise funds to be able to provide honoraria and appropriate compensation for part-time work for the community. Technical challenges - while we have technical expertise within our team, we are still exploring the open sources available for implementing the Forum mode and also for the indexing of papers and DOIs. In our proposal we have provided 3 options for this which the last one is our fall-back plan in case we hit technical roadblocks. Social challenges - change is not always easy. As community members transition from corporate social media to community-owned cooperatively-governed social media, we are going through some growing pains and our Social Working Group will need to work towards understanding and addressing the needs of the community. We’ve already noted that the community desires an algorithm for discovering relevant research and scientific discussions. While our Forum mode implementation could help with that, we’d need to make the integration of this mode with neuromatch.social appear seamless for widespread adoption.

Neglectedness: To the best of our knowledge, the Next Generation Internet grants like NGI-zero could fund this type of work but some of their calls require an association with Europe. Currently none of our working group members are located in Europe. We have not applied for funding for this specific work before mainly for 2 reasons: 1) We are currently in our bootstrapping phase with Neuromatch Inc. as our fiscal host. Our infrastructure costs are moderate and Neuromatch Inc. covers our monthly server costs (~100 USD/month) and some annual service-provider fees (~100 USD/annum). We are very grateful for their generous support during this phase where we are defining our bylaws, policies, and norms based on a cooperative consensus-based decision-making process. 2) Most of the day-to-day work is being done by working group members who are working completely on a voluntary basis on top of their paid jobs. Consequently it has been hard to find time for grant-writing and fund-raising while balancing our volunteer work to avoid burnout.

Success: Qualitatively we’d measure the success of this work based on feedback from the community and the kinds of discourse that take place in our forum. If the feedback about our organization and decision-making process is generally positive, we’d consider that as a success of the cooperative form of governance. If the discourse on neuromatch.social and the forum is largely productive leading to science outreach and communication of scientific knowledge, we’d consider that as a success! Quantitatively, we would measure the success of this work based on the adoption rate of the tools we plan to build and the number of members who join to become active members in the community with a high retention rate. We'd definitely count diversity within the community (gender, race, neurodiversity, seniority, nationality, etc) as a sign of success.

Total Budget: 25000

Affiliations: We are all members of neuromatch.social which is currently unincorporated, with Neuromatch Inc. as our fiscal host. We are applying as a team of individuals.

LMIE Carveout: Our community consists of members from all over the world with the top active non-English languages being German, French, Spanish, Chinese, Portuguese, and Turkish. We are aware that over-reliance on English hinders participation in science and scientific research itself (https://neuromatch.social/@manisha/110756186727892580). Manisha has contributed to the Hindi translation of the interface via Crowdin and recommends members to use browser-specific translation features. We hope that the fact that most of the Working Group members are themselves from marginalized communities (LGBTQ+, neurodiverse, non-native English speakers), provides encouragement that others from these communities are certainly an integral part of this vibrant community and shouldn’t hesitate to participate.

Team Skills: We have domain expertise in neuroscience, community moderation, cooperative governance, system administration, and software development. Our community consists of over 450 active members. We have partnered with a sibling instance synapse.cafe. From our team, Jonny and Manisha have been working on organizing scientific reform for the past few years. Jonny has been actively promoting the need for decentralized open infrastructure in (neuro)science (https://doi.org/10.48550/arXiv.2209.07493). Manisha brings in operations experience from her stint as COO at Neuromatch Academy. She’s been a moderator for various communities and in India she collaborates with the Echo Network - a social innovation platform working on embedding science within the community. Lina is a senior full-stack software developer with DevOps experience. Lina also brings to our team the best practices in software development that ensure equitable credit-assignment for all contributions to the source code. Lina and Jonny are also members of social.coop. We all have plenty of experience working openly and collaboratively across time-zones. We are a team that consists of trans, non-binary, and female working group members, and value representation of our often marginalized community within academia and society. Collectively we believe that this team and the neuromatch.social community has the technical expertise, social experience, and shared vision that are vital for the successful achievement of our goals.

Submission Number: 197