Cosmik: Collective Sensemaking Networks https://archive.cosmik.network Re-imagining how we make sense of science Wed, 05 Aug 2026 19:43:43 GMT https://validator.w3.org/feed/docs/rss2.html https://github.com/jpmonette/feed en All rights reserved <![CDATA[Cosmik Updates: April 2026]]> https://archive.cosmik.network/updates-april-2026 DJJjgVeRoYgLfZvIUKjF Mon, 11 May 2026 00:16:03 GMT Dear readers and Semblers,

Sorry we’re late with the April update, but a lot is blooming this spring for Cosmik and Semble! 🌱

Cosmik

  • We’ve been busy refining our program thesis as part of the Renaissance Philanthropy BiTS accelerator - we’ll have more to share next month.

  • We gave a talk at the STAMINA working group in the Complex Data Lab at McGill University in Montreal, view the recording here:

ATScience

  • The ATScience talk recordings are up! View them here

  • We’re collecting feedback from event participants - if you attended either in-person or via the livestream, please fill out the post-event survey!

  • Once we process the feedback we’ll share an event writeup

Semble

Connections

New knowledge sharing patterns

We’re continuing to notice interesting new knowledge sharing patterns on Semble.

Community

We continue to be blown away by the generativity of the atproto community.

What would you like to see next on Semble? We're always keen to hear feedback or suggestions so please be in touch!

Thanks again to all the awesome contributors, and see you in the Atmosphere!

Team Cosmik

]]> cosmik@newsletter.paragraph.com (Ronen Tamari) <![CDATA[Cosmik Updates: March 2026]]> https://archive.cosmik.network/updates-march-2026 g7e4mCZCmyiEywnUrgvn Wed, 08 Apr 2026 21:18:04 GMT Hello everyone,

March was an action-packed month!

AtmosphereConf and ATScience

We just got back from AtmosphereConf and ATScience in Vancouver, which we sponsored and have been busy co-organizing for the last few months. ATScience was fantastic - registration was maxed out at 125% and the day was full of great presentations, discussions and new connections sparking every which way. More details to follow in a separate post!

AtmosphereConf in general was a delight - the ecosystem is fun, vibrant and diverse. We also really appreciated the opportunity to engage with Semble users and hackers in real life. So many interesting leads and ideas to follow up on!

Check out our 2 talks here:

Cosmik

Semble

  • We’ve been hard at work on a new Connections feature we’re really excited about, inspired by our collaborations with the modular research and Discourse Graphs initiatives - more to follow soon!

  • Based on community feedback, we did some performance optimizations on card saving to make the core curation flow a lot snappier. Shout out to Roberto for his suggestion.

  • Following specific collections unlocks a new pattern we’re calling "faceted following" — instead of following someone wholesale, you can follow specific facets of their work. Faceted following also serves as a signal for curators: as a curator, you can see which facets people are gravitating toward. If a subtopic you've explored quietly keeps attracting followers, that's an interesting signal from your community to dig deeper into that subject.

Community

  • We really appreciate the thoughtful feedback we're getting from community members — it's directly shaping our development priorities and helping us make Semble work better for you. Shout out to Chris and Tyler for especially valuable input!

  • “Living libraries” are an exciting pattern emerging from the community: personal links pages or research literature reviews powered on the backend by Semble collections. For example, Nick Vincent's Data Counterfactuals site pulls from his curated Semble collections at build time, creating a reference shelf with dozens of papers that stays current as he adds and reorganizes content on Semble. What makes this powerful is that the content now lives on Semble rather than in static HTML files. Contributors can add references to a shared collection without pull requests or HTML editing. And because it's on Semble, you get all the affordances that come with it: anyone can follow the collection, get notified when new papers are added (coming soon!), and discover the content through Semble's network alongside everything else being curated there. The website becomes just one view of a living, collaborative resource. We're excited to see more of these.

  • One more wonderful emergent pattern has been collectively curated Semble collections that feed into newsletters. Shout out to Chris (again) who started at://news which is powering a Leaflet publication called Atmospheric weather report, with contributions from several regular contributors.

  • Tim Disney added Semble and Margin integration to their SkyReader RSS reader app!

  • Chive shipped a Semble intergration so you will be able to curate publications on Chive and mirror the collection to Semble.

That’s all for now, happy Spring, touch grass and may our community gardens (real and digital) bloom 🌻

Semble (@semble.so)

AtmosphereConf was absolutely epic but we really needed to get outdoors touch granite today 🏔️ Sembling in the wild @ronentk.me @pouriade.com @wesleyfinck.org

https://bsky.app
Semble (@semble.so)

]]> cosmik@newsletter.paragraph.com (Ronen Tamari) cosmik@newsletter.paragraph.com (Cosmik: Collective Sensemaking Networks) cosmik@newsletter.paragraph.com (Wesley Finck) <![CDATA[Cosmik Updates: February 2026 ]]> https://archive.cosmik.network/updates-february-2026 NtYna8dOCDi3Cfi0Fiwv Wed, 04 Mar 2026 16:54:46 GMT Hey everyone,

February was a fun and a busy month, we launched some important new Semble features including collaborative collections, collection following and interoperability with Margin (an ATProto bookmarking and annotation app). We also saw a lot of great community contributions, both by humans and AI agents(!)

Cosmik

We finalized the agenda for the ATProto Science event on March 27. We're really excited by the range and quality of proposals we received - the ATProto Science ecosystem is really popping! There's still time to register if you want to join us — we're almost at capacity but have a few slots open. See the agenda and registration details at https://atproto.science/events/atmosphere2026/

Agentic knowledge work

Lots of interesting developments at the intersection of ATProto and AI agents, in particular as support tools for sensemaking and annotation.

We see AI agents and social sensemaking tools as a powerful combination. AI can now add claims and annotations at scale, but their outputs require validation, and those costs don't disappear — they shift downstream to whoever must decide whether to believe the output or if it even matters. Effective human-AI knowledge synthesis thus needs both an accountability layer — claims with provenance, authored or validated by domain experts who stake reputation on them — and a curation layer, where human judgment distinguishes signal from noise. Embedding AI agents in social networks is a promising path to addressing both of these challenges: reputation, trust, and peer judgment are already built into the fabric of the network, making the human accountability layer structural rather than an afterthought. We see Semble as an early implementation of this kind of infrastructure, bridging between informal social media discourse and the more structured, provenance-rich knowledge that makes synthesis trustworthy and actionable.

Semble

Features

Community contributions

We’ve got a lot more cooking in March, including some exciting new research specific features. Stay tuned!

]]> cosmik@newsletter.paragraph.com (Ronen Tamari) cosmik@newsletter.paragraph.com (Cosmik: Collective Sensemaking Networks) cosmik@newsletter.paragraph.com (Wesley Finck) <![CDATA[What happened to Science Goodreads and how do we rebuild it? A 65 million dollar question (at least)]]> https://archive.cosmik.network/mendeley-story Yz04uBMDg4Dr51MJ9Vn1 Fri, 13 Feb 2026 16:14:27 GMT I've been telling the story of Mendeley at events for the last year, but never got around to writing it down. I found that while a lot of people are familiar with Mendeley, most are not familiar with the back story. The Mendeley story is emblematic of the dysfunctionality of the publishing system and the ways researchers are screwed over by extractive publishers, but it also contains the clues for a promising path forward. The story of Mendeley shines a light on the hidden value of the digital footprints left by researchers; a multi-billion dollar asset we collectively generate but don't control. Those digital footprints, and who controls them, are a key piece of the sustainability puzzle for research commons, which is why I think this story deserves to be much more widely known.

I first learned about the Mendeley story back in 2022 by asking a simple question that turned out to be a rabbit hole: why don’t we have something like Goodreads for science? I asked this back in 2022, but the question is still very much with me, and it keeps coming up with others as well; just last year, Sanje Horah posted this tweet:

Which I keep referencing, since it is so striking how much the idea still resonates with people: 8.4M views and thousands of likes and shares.

So why don’t we have something like Goodreads for science? TL;DR - we did. Mendeley was one of the first. Its story is representative, and roughly tracks the web’s evolutionary stages (e.g., web 1.0 → 2.0 → 3.0). Mendeley was a cool new “web 1.0” science goodreads, until it got eaten by Elsevier, a data-hungry publishing behemoth (Web 2.0). Web 3.0 is where we're at today: the decentralized dreams of the past few years mostly fizzled, but emerging cooperative federations and open social protocols are laying the groundwork for research commons that can't be acquired and dismantled the way Mendeley was.

That’s the TL;DR, but the details are worth diving into.

Web 1.0 - the good ol’ days

The basic facts of Mendeley’s story are available on Wikipedia. It was founded in 2007 by 3 German PhD students, Paul Foeckler, Victor Henning, and Jan Reichelt. Mendeley is a reference manager software used to manage, annotate and share research papers and to generate bibliographies for scholarly articles.

This may not seem particularly exciting, but by 2009 Mendeley had incorporated more advanced features such as personalized content recommendation and social networking. The  Guardian ranked it #6 in "Top 100 tech media companies", and it won a European Start-up of the Startup of the Year award in 2009:

“[Mendeley] describe themselves as the ‘Last.fm for research’. the startup essentially aims to enable academics to manage and share their research paper inventory and at the same time discover like-minded people and papers thanks to a recommendation and matching algorithm.”

So way back in 2009 we already had quite a popular Goodreads for science in the form of Mendeley.

Web 2.0 - When The Rebel Alliance Sells Out

The next chapter of Mendeley’s story is well told by a 2013 New Yorker piece by David Dobbs titled “When the Rebel Alliance Sells Out”. Mendeley “had become a sort of rebel-scientist icon for producing a software-and-paper-sharing service that threatened to disrupt scholarly publishing in the way that Napster and last.fm had disrupted the music industry a few years earlier.”

And more:

“Simply by searching, finding, reading, and sharing papers on Mendeley, researchers could rebel both personally and politically: they could enjoy the frisson of obtaining a paywalled paper without paying for it. And by unbundling the paper from the journal, as last.fm had let music fans unbundle songs from albums, they felt like they were kicking a brick out of the outmoded publishing infrastructure.”

But “as users became happier, publishers grew grumpier and investors less patient.” By the end of the summer of 2011, “Mendeley had over a million users who had uploaded over sixty million papers—the world’s single biggest repository of academic papers—and were sharing many of them.” The company was demonstrating a lot of potential, but it needed to start generating big profits to appease its investors. Mendeley needed an exit plan.

These conditions made Mendeley an attractive target for Elsevier, a publishing behemoth “infamous for restricting the flow of scientific information so it can sell research papers for as much as fifty dollars a piece, generating profit margins of thirty-six per cent and netting the company billions of dollars in revenue annually.”

In 2013 Elseveier acquired Mendeley for around $65 million. Elsevier, Dobbs writes, had two main reasons for purchasing Mendeley:

“One is to squash it—to destroy or coöpt an open-science icon that threatens its business model. Many critics fear that’s the case. The other reason is to possess the aggregated data that Mendeley’s users generate with all of their searching and sharing...Their use patterns reveal who is reading what, which papers are popular, what lines of research are surging, which disciplines and journals are crucial, and a lot of other extremely valuable information.”

After the acquisition, some still maintained hope “Mendeley would use their Elsevier backing to do something really fantastic”. Unfortunately, this was not to be, and instead Mendeley followed the well-trodden path of startups that have a “successful” exit for founders but end up “dying in the bosom of a larger company”.

For Elsevier, on the other hand, the acquisition of Mendeley was part of a broader transformation from a publisher into a data analytics company. Elsevier and other publishers recognized that the open access movement in science threatened their traditional closed access business model. They astutely realized that, in a world with open access to infinite information, the next big thing would be the data analytics tools needed to make sense of it all. By the late 2010s, publishers had transitioned into data analytics companies. Today, Elsevier no longer calls itself a publisher- they are providers of “advanced information and decision support to accelerate progress in science and healthcare worldwide.”

Jefferson Pooley, a researcher at the University of Pennsylvania, calls this new business “surveillance publishing,” highlighting the similarities with Shoshana Zuboff’s characterization of big tech platforms’ “surveillance capitalism” business model: “Elsevier and its peers are layering a second business on top of their legacy publishing operations, fueled by data extraction. They are packaging researcher behavior, gleaned from their digital platforms, into prediction products, which they sell back to universities and other clients.“

PlumX, an analytics company also acquired by Elsevier, lists Mendeley data under the “Captures” metric that measures how many times users bookmarked or saved research content. Interestingly, they note that captures specifically “can be a leading indicator of future citations” - hence the value of acquiring the data.

They note that:

“Technologies that encourage communication, sharing and other interaction with research output—leave “footprints” to show the way back to who is interested in the research and why.”

In business terms, data analytics translates into a lot of money: data from Mendeley, PlumX and other products fuels Elsevier’s “Databases and Tools” segment which now generates almost half their 3B GBP yearly revenue (2024).

Not bad for a $65M purchase back in 2013.

Web 3.0: when the Rebel Alliance strikes back?

Fast forward to today. We still don’t have Science Goodreads, but we’re starting to rebuild it, with Semble and a constellation of interoperating open social apps for research.

And this time around we have a unique window of opportunity to do it right:

Imagine what could be possible if we had researcher-stewarded ecosystems capable of generating even a fraction of the billion-dollar revenues extracted by platforms like Elsevier, but with that value reinvested back into the people and communities that created it. That's the vision we're building towards: not another startup that needs to find its exit, but a cooperative, self-sustaining commons that belongs to researchers, by design and by governance. This time, the Rebel Alliance won't need to sell out.

]]> cosmik@newsletter.paragraph.com (Ronen Tamari) <![CDATA[Cosmik January Updates]]> https://archive.cosmik.network/cosmik-january-updates Z9jr9nOJT6VidJZzaAHY Tue, 03 Feb 2026 21:34:30 GMT One of our New Year’s resolutions was to start sharing monthly updates, so here’s one for January!

Cosmik

  • We have been busy co-organizing ATScience 2026, a full-day exploration of ATProtocol for science, education, and open knowledge happening as a side event to AtmosphereConf 2026 in Vancouver on March 27, 2026. We were excited to see many great proposals (more than 20 total!) and are working on finalizing the agenda in the next week or two. We’d love to see you there if you can make it!

  • We launched the CAIROS initiative - a cooperative federation of open science projects working towards collectively stewarded research commons, co-founded with Discourse Graphs. CAIROS aims to forge new paths for mission-driven open science ventures by combining modular research, open social networks, and cooperative governance.

  • We've been gearing up for the MIRA workshop (Modular Interoperable Research Attribution), a five-day workshop in Ireland this summer. MIRA is organized by Discourse Graphs and other collaborators to design and prototype interoperable frameworks for modular research attribution, bringing together researchers, designers, engineers, and ecosystem specialists to build next generation scientific communication and collaboration systems. We're particularly excited by MIRA's goal of integrating open social protocols like ATProtocol directly into research workflows - bringing social sensemaking and collaborative knowledge building into the heart of the research process itself, rather than treating them as disconnected afterthoughts.

Semble

On Our Radar

  • So many great ATProto Science projects! Check them out in the aforementioned Semble collection.

  • From Albums to Streams: How Modularity Changes Systems - by Rowan Cockett from Continuous Science Foundation on unbundling research into modular knowledge: “In a physical world of .. printed journals, bundling made sense. In a digital world, we are increasingly constrained by boundaries that no longer make sense. In a world of AI, what does the bundle prevent us from seeing?”

  • Context Widows - Great piece around the question “can LLMs do science” and a useful reframing of the discussion: “The better question is whether LLMs can be usefully integrated in the processes by which scientific knowledge gets made, validated, and disseminated.”


]]> cosmik@newsletter.paragraph.com (Ronen Tamari) cosmik@newsletter.paragraph.com (Cosmik: Collective Sensemaking Networks) cosmik@newsletter.paragraph.com (Wesley Finck) <![CDATA[Cosmik New Year: 2025 in Review, and Looking Forward to 2026]]> https://archive.cosmik.network/cosmik-25-26 YgDsZFhe3wbR62w5BZCo Wed, 14 Jan 2026 01:19:38 GMT As we begin 2026, we want to share some thoughts for the upcoming year and also take a moment to reflect on 2025 which has been an exciting year for Cosmik.

Launching Semble

The headline for 2025 is the November alpha launch of Semble, our social "micro-knowledge" network for researchers. When we announced our grant funding from Open Philanthropy and Astera Institute back in August, we described Semble as Are.na meets “Goodreads for research”: a place where researchers can create, organize, and share knowledge in ways that complement traditional feed-centered social platforms. Five months later, that vision is live.

The response has been encouraging. In the two months since our launch, 150 users joined Semble, collectively creating over 5,000 cards. Sembling is now a thing:

Torsten Goerke (@tgoerke.bsky.social)

Reading this forum post on open protocols and music publishing. Thank you @hilk.eu for sembling it. https://the.socialmusic.network/t/protocols-not-platforms-a-technological-approach-to-free-speech/382

https://bsky.app

We’re excited both by users seeing value in Semble as it currently is, as well as users sharing in and expanding on our vision of what it could become. As our friend and collaborator Erlend Sogge Heggen observed,

Semble is the most radical atproto app around, as it represents a departure from entertainment-feeds in favor of attainment-feeds (learning), which fundamentally changes the game of social media as we know it.

We've also released a public API, and we're already seeing community developers build on top of Semble. Projects include Raycast extensions, website integrations, and even explorations of how Semble could connect with tools like Bluesky bots (see the full list on Semble).

What We've Learned

Launching Semble has given us early evidence for a hypothesis we've long been excited to test: that the right interfaces can harness valuable "creative exhaust"—the thoughts, connections, and references that researchers generate but rarely share publicly. Semble also makes space for new kinds of knowledge creators—for example, we are noticing people who are quiet on Bluesky but who curate fascinating content on Semble.

We've also learned that ATProto is a wonderful ecosystem to be building in, and fertile ground for the kind of community-driven development we want to foster. ATProto’s emphasis on interoperability and user ownership aligns well with our vision, and there's growing institutional interest (such as the EuroSky initiative) that could help this ecosystem really come into its own.

That said, our alpha is still quite early with a lot of room for improvement. Users are finding value in the current version, but we know there is significant potential for expansion, both on the creation side (for example, integrations making it easier for users to add content to Semble) and the discovery side (for example by surfacing relevant connections and helping users’ content get seen).

Building the Ecosystem

Alongside product development, we've invested significantly in ecosystem building—connecting with aligned organizations and helping foster communities that share our vision for the next generation of collaborative research infrastructure.

Some highlights:

  • We launched the ATProto Science community and are now organizing the first ATProto for Science workshop, scheduled for Vancouver in March 2026. This will bring together researchers, developers, and institutions interested in building research applications on open protocols.

  • We initiated CAIROS (Collective Augmented Intelligence for Research & Open Science) in partnership with the Discourse Graphs team. This initiative explores how research communities might collectively own and govern their knowledge infrastructure—a crucial piece of the puzzle for long-term resilience and sustainability.

  • We joined the Continuous Science Foundation's Modular Peer Review working group and have been participating in Renaissance Philanthropy's Future of Science Curation conversations.

These connections are helping us weave together something bigger than any single product could be. As we recently put it: the killer app in open social protocols is a killer ecosystem.

(source)

Events and Presentations

This year took us to many events, where we've had the opportunity to share our vision and learn from others working on similar problems. Highlights include:

Looking Ahead to 2026

The coming year will focus on four priorities.

First, we'll continue to build out Semble's core functionality by improving personalization and discovery, making collections more collaborative, and AI enhancements. We'll also add more research-specific features to Semble, such as citation metadata, annotation tools, and relations between cards. Throughout this process, we're committed to building in the open and actively incorporating feedback from the community to shape our development priorities.

Second, we’ll develop integrations with other research platforms as well as ATproto tools. These integrations will help Semble meet researchers where they are, support novel workflows, and enable capabilities that no single app could offer alone.

We’re eyeing a rapidly growing list of exciting integrations, including:

  • Discourse Graphs for knowledge synthesis and structured micropublishing

  • Seams.so and margins.at for atproto annotations

  • Lanyards and ORCID for credentials

  • Roomy.space for empowering link sharing in group chats

  • Leaflet for weaving Semble into long form writing

  • Chive for adding Semble data and affordances to preprints on ATProto

  • Bluesky integrations for tapping into the rich discussions already happening there

  • Lea (the exciting new Bluesky research client) for connecting Semble more directly with Science Bluesky

Third, we’ll be investing in IRL gatherings for high bandwidth sensemaking and community building. We will participate in and co-organize events with ecosystem partners, including ATScience 2026 in March and the Modular Interoperable Research Attribution (MIRA) workshop in June.

Finally, we'll be ramping up our efforts around sustainability. We’ll explore novel philanthropic funding models like Focused Research Organizations and Coordinated Research Programs. Through CAIROS we will develop our longer-term vision of self-reliant and self-sustaining research commons—potentially through individual and institutional memberships, data trusts, or hybrid models like Mozilla's.

Thank You

None of this work would be possible without the support of Open Philanthropy and the Astera Institute, whose generous grants have allowed us to pursue this vision. We're also deeply grateful to our early users, advisors and collaborators who have shaped our thinking and helped us understand what the community actually needs.

2025 was the year Semble went from concept to working product. 2026 will be about making Semble shine—and weaving it into a vibrant ecosystem for collective intelligence.

]]> cosmik@newsletter.paragraph.com (Ronen Tamari) cosmik@newsletter.paragraph.com (Cosmik: Collective Sensemaking Networks) cosmik@newsletter.paragraph.com (Wesley Finck) <![CDATA[ Introducing CAIROS: Collective Augmented Intelligence for Research and Open Science]]> https://archive.cosmik.network/introducing-cairos MlfTIA41V0zEslG4bhsc Fri, 09 Jan 2026 16:11:46 GMT We’re excited to open 2026 with two fresh new initiatives! The first, ATProto Science, we announced just before the holidays.

The second and the subject of this post is CAIROS: Collective Augmented Intelligence for Research and Open Science.

CAIROS is a cooperative federation of open science projects working towards collectively stewarded research commons. The founding members are Cosmik Network and Discourse Graphs. The back story is that we, along with many mission-driven open science ventures, are searching for cooperative alternatives that more naturally align with the principles of open research infrastructure. Traditional pathways are often limited to precarious grants, restrictive VC funding, or acquisition by commercial large publishers.

We hope CAIROS will help forge new paths, empowering us to join forces in building a more collaborative science together.

We presented CAIROS in December at a workshop on "Standards, Stewardship, and Solidarity: Building Flourishing Science Commons” as part of AGU 2025 (American Geophysical Union):

Continue below for an overview, or see the publication with the slides and abstract.

Why now?

Scientific communities are straining under mounting challenges—knowledge fragmentation, outdated publication processes, institutional erosion and funding cuts that threaten our capacity to address urgent global problems. Current incentives are driving scarcity and individual competition rather than the open knowledge sharing and large scale collaboration we need in order to overcome the “science polycrisis”.

The promise and peril of tech

Transformative technologies like AI and open social networks stand to play a pivotal role in the next phase of science: on one hand, they create unprecedented possibilities for addressing information overload and empowering collaborative research.

On the other hand, they also bring unprecedented risks of “slop science”, power centralization, and community fragmentation.

These challenges and opportunities call for new ways of organizing research, fundamentally reimagining how it's created, who can participate, how it is disseminated, and how the value generated from research is distributed

CAIROS

Kairos is an ancient Greek word meaning the right or critical moment for words, action or movement.

CAIROS is our attempt to meet this moment, by integrating three innovations at the metascience frontier. The first two will be familiar to readers of the blog, the third is more CAIROS-specific.

1. Modular “permaculture” research for human & machine readable knowledge

Modular research, exemplified by initiatives like Continuous Science Foundation, Nanopublications and Discourse Graphs, is about unpacking the research process into machine readable, composable “knowledge units” like questions, claims, and evidence. Where traditional research has turned into a monoculture that recognizes only publications, modular research unlocks a new kind of “permaculture science” which recognizes diverse contributions, empowers citizen science and enables grassroots, distributed large-scale collaboration.

2. Bridging social media and scientific discourse

Modular research is great in theory, but a central challenge remains how to gain widespread adoption. Science social media provides a key piece to the incentive puzzle. Researchers on social media, when they share questions, recommendations or critiques, are engaging in modular research already, they just don’t know it yet. The problem is that current social platforms are not designed for science, so those insights too often get lost in the noise or enclosed by data hungry platforms.

The good news is that we’re at a unique moment where open social networks like Bluesky are seeing huge popularity gains with scientists. And we can go even further than that - open platforms are built on open protocols like ActivityPub and ATProtocol, which are inherently extensible and enable the creation of whole new kinds of social media tailored for research. When effectively federated, these protocols and tools have the potential to unlock new forms of participation and knowledge sharing that doesn’t happen on traditional platforms.

Examples include our Semble and other ATProto Science projects, and Bonfire’s Open Science Network on the Fediverse.

3. Collectively stewarded research commons

We recognize that systemic change will require organizational and not just technical innovations.

We take inspiration from works like Samuel Moore’s excellent book Publishing Beyond the Market. Moore makes the case for commons as a third way beyond purely commercial, governmental or philanthropic models - a mode of production grounded in collective self-organization and self-reliance.

We view cooperative federation as a promising path towards implementing open and sustainable research commons. Cooperative approaches empower democratic governance, prioritize collective progress over individual competition, and unlock new forms of equitable data stewardship such as data cooperatives.

Collective Augmented Intelligence

The idea of collective intelligence is central to our initiative. AI-centric approaches focus on wholesale task automation and risk automating humans out of the loop. In contrast, collective intelligence systems are designed for virtuous cycles where AI and humans improve each other recursively, thus preserving and empowering human agency

To give a sense of what we mean by virtuous cycles - imagine new social networks that incentivize participation and elicit authentic knowledge sharing, tools that structure this thinking into modular units, and hybrid human-AI systems powered by that data to support next generation curation, discovery and synthesis:

Cooperative governance at the center is what keeps the cycle spinning and humans in the loop. It maintains human oversight across all components and ensures value is reinvested back into the ecosystem. This keeps the entire cycle aligned with research communities rather than extractive interests.

Collective intelligence isn't just about improving collective knowledge work: as explored by the Collective Intelligence Project, CI also enables and requires transformative new business models.

Connect with us

If you’re interested in joining the movement to create collectively owned research commons, please reach out at https://forms.cairos.network/connect or hello@cairos.network

Whether you're building open science projects or communities, contributing as a researcher, or funders exploring how to support this transformation - we’d love to hear from you!


]]> cosmik@newsletter.paragraph.com (Cosmik: Collective Sensemaking Networks) cosmik@newsletter.paragraph.com (Ronen Tamari) <![CDATA[Introducing ATProto Science]]> https://archive.cosmik.network/introducing-atproto-science 2i57zFhkoZEVwH6Ov7vv Tue, 23 Dec 2025 20:41:58 GMT Just before you log off for the holidays, we’re excited to share news about ATProto Science, a new initiative we’re launching that we’ve been partnering on with collaborators in the ATProto ecosystem.

ATProto Science (@atproto.science)

Announcing atproto.science! #ATproto has huge potential for supporting new ways of doing research, so we started @atproto.science to help bring together researchers, build community, and explore what's possible when science meets cutting-edge open social network protocols! 🧪 > http://atproto.science

https://bsky.app

ATProto for Science (or ATScience for short) is an initiative to connect and support researchers, developers, and communities exploring how ATProto can serve research, education, and open knowledge work.

Cosmik is also co-organizing and sponsoring the first ATProto for Science workshop! It will be a full-day, in-person gathering (March 27, 2026 in Vancouver, BC) during the main ATProto conference, AtmosphereConf 2026 (March 26-29). See below for registration details and a call for proposals.

Why Now?

ATProto has huge potential for supporting new ways of doing research, and this initiative is intended to help catalyze the efforts that will get us there!

Research communities face increasing fragmentation, platform fragility, and funding uncertainty. Science social media has long suffered from commercial platforms that come and go, change their policies, and in some cases actively extract value from the communities that sustain them. Meanwhile, government cuts to science funding make resilient, self-governed infrastructure more critical than ever.

ATProto offers the promise of research networks that communities can meaningfully build, control, adapt, and maintain themselves.

What Is ATScience?

ATScience is an exciting nascent space with a lot of emergent projects and activity, including:

  • Researchers using Bluesky for science communication and collaboration

  • Lively discussions around new visions for science social media. For a recent example, polarized discourse around AI on Bluesky triggered interesting conversations on new moderation affordances to protect researchers from harassment.

  • Builders creating research tools tools on Bluesky and ATProto: credentials, feeds, review and curation (such as our semble.so!)

These efforts exist, but they are still scattered. ATScience aims to weave them together into an ecosystem, creating spaces for connection, collaboration, and shared learning.

Groups like Blacksky and Northsky are exploring community governed social media infrastructure. Could we do the same for science social media?

Who Is ATProto Science For?

We're thinking broadly about "science": traditional academic research, but also community science, independent research, education, self-teaching, and any form of rigorous knowledge work.

If you're:

  • Building research tools or infrastructure on ATProto

  • Conducting research on or using Bluesky and ATProto platforms

  • Reimagining scholarly publishing and communication in the age of AI and open social protocols

  • Interested in how decentralized protocols can support new forms of research and knowledge work

  • Curious about alternatives to extractive academic platforms

...this is for you!

Attend ATScience 2026

Our first major event is a full-day workshop at ATmosphereConf 2026 in Vancouver, BC on March 27, 2026! For event details and registration:

https://atproto.science/events/atmosphere2026/

Present at ATScience 2026

Want to present your work, lead a discussion, or demo your project? We’re looking for contributors across all areas of science on ATProto or open social media more broadly, from academic research to new knowledge tools or grassroots citizen science. Submit your proposal here:

https://forms.atproto.science/atscience26-proposals

Get Involved

We're figuring this out together. If you're working on something in this space, thinking about these questions, or just curious:

Happy holidays!

]]> cosmik@newsletter.paragraph.com (Ronen Tamari) <![CDATA[Cosmik December Lab Notes: 'Tis the season for sharing your best finds, and other Semble updates]]> https://archive.cosmik.network/gems-of-2025 8D3kanEyAYcj1XJd4gVl Tue, 16 Dec 2025 15:57:28 GMT Happy December everyone!

‘Tis the season for sharing your best finds 🤓📚☕❄

To celebrate the end of the year and get our holiday reading lists in shape, we're inviting you to make a “💎 Top picks of 2025” list on Semble.so  with books, articles, videos, tools, or anything worth passing on!

We’ll feature your picks in a special end-of-year “💎 Gems of 2025” feed on Semble. To get started, just login to Semble and follow the directions on the Home screen to create your list or view others:

Or see our Bluesky post for more details.

New feature: Notifications!

We added basic notifications, to help spread good attention sharing karma ✨

When someone collects a card via your card, you will now receive a notification:

Provenance Field in Lexicons

Part of these notifications is tracking where someone saved a card from, so we've updating our lexicons to include a provenance section with a via field which is a strong ref to the original card record. Here's an example of what this looks like:

{
  "type": "URL",
  "$type": "network.cosmik.card",
  "content": {
    "url": "<https://lab.leaflet.pub/3m7py65pswk2e>",
    "$type": "network.cosmik.card#urlContent",
    "metadata": {
      "type": "link",
      "$type": "network.cosmik.card#urlMetadata",
      "title": "Atmospheric @-mentions! - Leaflet Lab Notes",
      "imageUrl": "<https://leaflet.pub/lish/did%253Aplc%253Abtxrwcaeyodrap5mnjw2fvmz/3lppk75kw7k26/3m7py65pswk2e/opengraph-image?6815eb61f733905a>",
      "description": "Lab Notes 017: mention Bluesky accounts, Leaflet publications, specific posts and more to come — an interface for links across the atmosphere",
      "retrievedAt": "2025-12-12T03:02:17.305Z"
    }
  },
  "createdAt": "2025-12-12T03:02:17.305Z",
  "provenance": {
    "via": {
      "cid": "bafyreigyya3ofkpxtpqhqtqlcny3vyvcaqfpfnzn4pikk6elt2ckvx5qcq",
      "uri": "at://did:plc:rtf3bjc3w2yn4syxtm4r7jt2/network.cosmik.card/3m7q5sjs2gs2d"
    },
    "$type": "network.cosmik.defs#provenance"
  }
}

If you include this field, the author of the referring card will get a notification that you saved their card. We’ll update the Semble PDS Client to include this field soon.

Recent Collections

Some of you have asked for this - when you add a new card you will now see your recent collections in the “Add to collections” section:

Community highlights

  • @laurenshof.online created the latest ATmosphere Report with a curated collection of interesting atproto articles from November. What's cool about this is that the curated links are now on-protocol as a Semble collection, with a custom page that lets you view them alongside ATmosphere mentions, on Semble, Leaflet and Bluesky. Check it out at https://traverse.connectedplaces.online/!

Laurens (@laurenshof.online)

New: ATmosphere Report #145, with collections of links to interesting atproto articles for November This time, the links are on-protocol, as a @cosmik.network collection, with a page that lets you follow the conversations on @leaflet.pub and @bsky.app traverse.connectedplaces.online

https://bsky.app

]]> cosmik@newsletter.paragraph.com (Ronen Tamari) cosmik@newsletter.paragraph.com (Cosmik: Collective Sensemaking Networks) cosmik@newsletter.paragraph.com (Wesley Finck) <![CDATA[Cosmik Lab Notes 002: Semble Updates - Community Tinkering, Basic API, and New Discovery Features]]> https://archive.cosmik.network/cosmik-lab-notes-002-semble-updates n3wwc9P3Ks40la43jUjt Fri, 21 Nov 2025 21:57:48 GMT Happy Friday everyone,

First of all - wow! Just two weeks have passed since we released the open alpha version of Semble, and it’s been very exciting to see the engagement!

Feeling the ACI (Augmented Collective Intelligence)

We’re already seeing a steady stream of interesting curations, and we’re starting to experience the magic of more intentional online attention sharing.

(source)

Community hacking on Semble

One development we are particularly excited by is that people are already hacking on Semble to add various extensions and integrations! So much is going on that we created a Semble collection to track all those projects:

Semble Community Tinkering (by Cosmik Network) - Semble

Community projects extending Semble in interesting ways

https://semble.so

We also recently merged our very first community PR! Shout out to @vielle.dev for making that happen!

PDS as Universal API

We recently shipped a firehose subscriber which listens for repository commits on network.cosmik.* records and automatically reflects these changes in Semble’s AppView:

What this means is that you can use your own PDS as a defacto Semble API (until we ship a more comprehensive AppView API). This is one of the really cools things about building on ATProto.

(source)

Or as Dan Abramov puts it in Open Social, "The protocol is the API.” Exciting stuff!

This is especially exciting for us because a few people have already asked us if it’s possible to support integrations with other bookmarking tools, for example to import bookmarks from apps like Raindrop. So to all the SenseMakers out there who like hacking on their knowledge tools — this is our first step in making that easier.

We’ve also published a lightweight PDS client purpose-built for working with Semble records in your repository. This makes it even easier to work with your PDS as a Semble API. Read more about how to get started with that in our docs:

Getting Started with the Semble PDS Client | Semble Docs | Cosmik

We'll be shipping a more comprehensive API in the future. Meanwhile, this is a simple and lightweight PDS client for creating, updating, and deleting Semble records directly on your PDS. Please share any feedback about bugs or ways to improve it (or feel free to make a PR directly to the package).

https://docs.cosmik.network

Similar Cards

The Semble page got a new tab — “Similar Cards”!

The “Similar Cards” tab surfaces semantically similar URLs from across the network, adding a whole new discovery layer within the app.

Bluesky post view

Finally, we also have a special new view for Bluesky posts (identifiable by the 🦋) that better preserves the original structure and content:

Happy assembling!

We’re continuously blown away by the creativity of the broader ATProto community and are grateful for all those that have taken an early interest in Semble. We can’t wait to see what we create together 🫶

— Cosmik Team

]]> cosmik@newsletter.paragraph.com (Ronen Tamari) cosmik@newsletter.paragraph.com (Cosmik: Collective Sensemaking Networks) cosmik@newsletter.paragraph.com (Wesley Finck) <![CDATA[Cosmik Lab Notes 001: Semble Alpha]]> https://archive.cosmik.network/lab-notes-001-semble-alpha YOKmMl5lmaMFyw6PSldt Mon, 10 Nov 2025 14:37:37 GMT We’re excited to start Cosmik Lab Notes, where we’ll be working in the open and sharing thoughts as we build Semble and the wider Cosmik Network.

(also posted on Leaflet)

⛳ Where we are: Semble Alpha!

Semble is in open alpha now and we’d love for you to try it out! -> semble.so

As observed by one of our early users, alpha is "a bit bare bones...but there are tons of opportunities with this"

Missino (@onedeuxtriseigo.nullpo.dev)

I am ecstatic for this. I'm gonna start using this for tracking reading, research, and references for various projects. It's a bit bare bones while it's in alpha but there are tons of opportunities with this. [contains quote post or other embedded content]

https://bsky.app

Semble is modeled after social knowledge tools like Are.na and Sublime. Right now you can collect links, organize them into collections, and see what others on the network are doing. Check out our other inspirations in this Semble collection:

Semble Product Inspiration

Products we take design, functionality or organizational inspiration from for semble

https://semble.so

To use Semble you currently need a Bluesky account, but in the future we'll provide an independent Cosmik ID.

To get started:

  • Login with your Bluesky account

  • Add a new card either from scratch (button on bottom left) or by clicking on the “+” symbol on someone else’s card. Currently, cards are simply links with an optional note attached to them.

  • You can also optionally create new collections and add cards to them.

  • Add our bookmarklet to your browser to quickly open any webpage in Semble.

  • Check out this guide for mobile setup

Semble page

Clicking on a card will take you to the Semble page: a page aggregating all of the activity around the link in that card, such as collections it's featured in, and comments about the link. So each link shared on the network will have a corresponding Semble page, which you can also find at https://semble.so/url?id=/<link-url>[1]. So for example, https://semble.so/url?id=https://cosmik.network will aggregate the activity around https://cosmik.network/. In the future we plan for the Semble page to aggregate activity from other ATProto apps and beyond.

🗺 Where we’re headed

We're at early stages but we’re already learning a lot from our alpha community. We have a lot of directions we're eager to explore, including:

  • Bluesky social graph integration to personalize your Explore feed

  • Collaborative collections

  • Bluesky, Leaflet and other atproto app integrations. Read more about our thinking around this here.

  • API and developer tools for custom integrations, for example importing data from other bookmarking apps.

  • Getting a broader set of link metadata (such as bibliographic citation data, DOIs, and so on)

  • Research-specific "emojis” for higher signal curation

  • Inline highlights and annotations

  • Semantic search to surface similar links curated by the network

… And more to come in future posts! Check out our public roadmap here.

Finally, we’re excited to evolve Semble with you all, so please share thoughts in the comments, leave an issue on Github or Tangled, or reach out!

— Cosmik Team

[1] We also support the simpler url redirect structure https://semble.so/url/https://cosmik.network, inspired by the excellent henry.ink

]]> cosmik@newsletter.paragraph.com (Ronen Tamari) cosmik@newsletter.paragraph.com (Cosmik: Collective Sensemaking Networks) cosmik@newsletter.paragraph.com (Wesley Finck) <![CDATA[The Fork in the Road: AI Horseless Carriages or Collective Intelligence for Science]]> https://archive.cosmik.network/collective-intelligence-for-science nFfhZgrquO95aIUSF9Sq Wed, 15 Oct 2025 20:20:25 GMT Last month I attended an in-person gathering of scientists, technologists and field-builders at the intersection of AI and metascience. This post is an extended version of a talk I gave there. It contains some themes you might be familiar with if you’ve read previous posts, but hopefully with some interesting new frames. The prompt for the talk was an invitation to explore one of the main discussion topics at the event - metascience threats and opportunities posed by AI acceleration.

To talk about threats and opportunities I’ll introduce the AI Horseless Carriage metaphor, from a blog by a developer named Pete Koomin. “Horseless carriage” refers to the early motor car designs that borrowed heavily from the horse-drawn carriages that preceded them.

Horseless Carriage (https://en.wikipedia.org/wiki/Horseless_carriage)

These early experiments swapped engines instead of horses, rather than rethinking vehicle design from the ground up. The AI horseless carriage story is at its core a warning: when a revolutionary new technology like AI is invented, the first tools built with it risk failure because they mimic the old way of doing things.

What are the “horse-carriage” risks in AI for science? And what is the “car” that we could be designing for instead?

I would argue that the AI horseless carriage is plugging AI into the existing publishing system. The result, taken to its somewhat absurd but not unlikely endpoint, looks like a loop of AI peer review on AI generated papers. I call this scenario a horseless carriage since papers and peer review in their current form are old ways of doing science that we already know are obsolete.

To be clear, AI automation is highly valuable for well-scoped tasks such as screening literature, extracting data, or running standardized protocols. The risk isn't automation itself, but automating broken systems. Such wholesale automation risks devolving science into a meaningless “doom loop”, with AI acceleration creating a kind of centrifugal force that deskills humans and pushes them out of the research loop (see figure below). We're seeing similar doom loops emerge in education, where AI-generated assignments meet AI grading systems, progressively removing students from meaningful learning while appearing to increase efficiency. This kind of automation is pernicious since in many respects it is the path of least resistance: it requires minimal institutional change and aligns with big AI companies' push to sell automation at scale. It also makes certain myopic economic sense, as cuts to science funding render human labor more scarce and incentivize institutions to look to automation as an imperfect but cheaper substitute.

But there is a promising alternative path, and what I would argue is the "car" we should be building: collective intelligence (CI) for science. In its broadest form, CI can be seen as “a collaborative approach to problem-solving, typically supported by technological tools, which allows for real-time co-ordination and mobilization of knowledge that is distributed among many individuals.”

The key difference between these paths lies in how AI relates to humans. In the horseless carriage scenario, AI automates tasks, deskilling researchers and pushing them to the periphery. In collective intelligence systems, AI instead amplifies and connects human thinking: facilitating collaboration, aggregating distributed knowledge, and helping groups coordinate at scales previously impossible. The focus shifts from automation to augmentation, and from substitution to synergy. In the horseless carriage scenario, each advance in AI capability tends to further marginalize human participation. In CI systems, improvements benefit both: better AI tools enable richer human collaboration, while more human participation generates better training data and clearer signals for AI. A particularly promising approach called Hybrid Collective Intelligence specifically designs for these virtuous cycles, where AI and humans improve each other recursively.

Two pathways of AI acceleration deskilling (left) vs upskilling through collective intelligence systems (right). Inspired by this tweet.

Achieving this kind of human-AI synergy would be transformative, and, it could be argued, even necessary for science. It would be a car, not a horseless carriage. But it will also be challenging, requiring us to rethink science communication infrastructure from the ground up.

Applying CI theory to the practice of research

Rethinking science communication is long overdue, and as Prachee Avasthi recently wrote, we should be treating it as a research problem in and of itself. The good news is that CI can help us there too! The key insight is that we can view science, and in particular science communication, as a CI system, albeit one using very outdated technology and practices. Perhaps we can apply insights from collective intelligence theory to improve the practice of science itself? This question has received surprisingly limited attention to date, but one notable work discussing it is the excellent Science Communication as a Collective Intelligence Endeavor. The piece identifies several key features that distinguish collective intelligence systems and explores how these could transform science communication. Four features are particularly relevant to our discussion:

  1. Structured aggregation of distributed knowledge, data and discourse: enabling synthesis at scales impossible for individual researchers.

  2. Increased participation by both researchers and the broader public: democratizing contribution while maintaining epistemic standards.

  3. Increased diversity of contributions and participants: bringing multiple perspectives to bear on complex problems.

  4. Increased responsiveness through rapid feedback cycles and the ability to respond to updates and new information in real-time.

In the time since the paper was published, new publishing tools and recent AI developments are combining to drive renewed interest in this problem space. Three metascience trends are now converging to open a unique window of opportunity for translating CI theory into practice.

Three converging metascience trends

Circle 1: Modular research

Already in a 1963 talk, Nobel laureate Peter Medawar argued that "the scientific paper is a fraud”, since "it misrepresents the processes of thought that […] gave rise to the work that is described in the paper." Modular research is an exciting trend in science towards publishing those processes of thought more faithfully. New tools like Nanopublications, Discourse Graphs and our Semble enable publishing precise and machine readable representations of atomic knowledge units such as hypotheses, observations, questions, reviews or even reading recommendations. Unbundling research into these connected and remixable units unlocks larger scale distributed research, similar to how the structured and modular nature of code facilitates the coordination of large scale open source software projects.

CI dimensions supported: structured aggregation, increased diversity and participation (1,2,3)

Circle 2: AI for science

The second circle is AI. Medawar’s insight has aged well - more than 60 years later, in the age of AI, we’re realizing that we can’t train AI models to think properly if we don’t have faithful representations of our actual thought processes. As Astera co-founder Seemay Chou recently wrote - “Automated AI agents can’t learn scientific reasoning based on a journal record that presents only the glossy parts of the intellectual process, and not actual human reasoning. We are shooting ourselves in the foot.” The success of AI coding assistants is a hugely impressive demonstration of the potential at the intersection of AI and modular research - what is code if not modular, structured reasoning? Training on large scale code data has unlocked powerful new forms of AI reasoning. Now “all” we need is Github, not just for code, but for modular, structured knowledge.

CI dimensions supported: scalable knowledge aggregation, increased responsiveness (1,4)

Circle 3: Science social media

So how do we go beyond the glossy thinking of papers to find actual human reasoning in the wild? How do we get people to write less papers and author more nanopublications? This brings us to the third and final circle, social media. Science social media is one of the most interesting new frontiers of scientific discourse. Michael Nielsen wrote about it in Reinventing discovery: the new era of networked science: “online tools are transforming the way scientists make discoveries. These tools are cognitive tools, actively amplifying our collective intelligence, making us smarter and so better able to solve the toughest scientific problems”. Science Twitter was indispensable to me during my PhD, especially since everything was online due to COVID. It was where I could get in the heads of experts in a way that would have been impossible just by reading their papers. Yes, science social media is clouded by hype and distractions and self promotion. But beyond those clouds, new forms of research are taking shape. As we’ve previously written, social media provides a key piece of the incentive puzzle for broader adoption of modular publishing practices: researchers are literally already nanopublishing there already, they just don’t know it yet. While social networks still lack the structure provided by modular research tools and the scalable aggregation capacities provided by AI, new open social protocols like ATProto are widely used by researchers and offer an exciting space for experimenting with new kinds of science social media (drop in and say hi here!).

CI dimensions supported: increased participation, responsiveness and diversity (2,3,4)

Note: These mappings are approximate and are meant to highlight each trend's primary contributions to CI. Also, each trend supports multiple CI dimensions, often in complementary ways.

Building the car: Collective Intelligence for Science

The synergy between these circles is why we think the stars are aligning for work on collective intelligence for science. While each trend has primary strengths—modular research for structured knowledge, AI for scalability, social media for participation—their real power lies in combination. This, we propose, is the car. Imagine social networks that incentivize and elicit authentic knowledge sharing, tools that structure this thinking into modular units, and AI systems powered by that data to support next generation discovery and synthesis. A cooperative human-AI Github for knowledge.

Instead of the AI horseless carriage where science drowns in AI slop papers, let’s build systems that keep humans firmly in the loop as essential participants.

And instead of the centrifugal force that pushes humans out of the loop, let’s create centripetal forces that draw humans together, using modular research, AI, and social networks to enable new forms of collaboration and discovery.

]]> cosmik@newsletter.paragraph.com (Ronen Tamari) <![CDATA[Cosmik awarded $1M grant from Open Philanthropy and Astera Institute for new social knowledge network for researchers]]> https://archive.cosmik.network/cosmik-grant-open-phil-astera PCiIBW4gOxNHsNxDmggG Fri, 15 Aug 2025 17:07:54 GMT We’re proud to announce that Cosmik has been awarded a total of $1M in grant funding from Open Philanthropy and the Astera Institute! These generous 2-year grants will support the next chapter of Cosmik and our work at the intersection of next-generation collaborative knowledge tools, AI and science social media.

Specifically, the grants will support the development of Semble, a micro-knowledge sharing and discovery network for researchers on Bluesky/ATProto. Semble is rooted in the idea of the ensemble, where diverse contributions harmonize to create richer, more nuanced understanding than any single perspective could achieve alone.

Think of Are.na combined with Goodreads for research, re-imagined for decentralized social networks and user-owned data.

Some early designs for Semble, a new social knowledge network for researchers. Think Are.na combined with Science Goodreads, re-imagined for decentralized social networks and user-owned data.

Why now?

In March we wrote that the stars are aligning for new kinds of social media for researchers, and that opportunity has only grown stronger since.

Bluesky has become a new hub for science social media, but we’ll need more than just Twitter clones

Altmetric recently reported that Bluesky now regularly eclipses Twitter in terms of the amount of research being shared, despite being much smaller.

Altmetric (@altmetric.com)

The trend is still in Bluesky's favour. X is still larger but very regularly Bluesky easily eclipses it in terms of research being shared. Linda Yaccarino claimed X had 600m users in April. Bluesky has 36m accounts created in total. [contains quote post or other embedded content]

https://bsky.app

The importance of this trend cannot be overstated. But we also need more than just Twitter clones - to really flourish, we need new social media experiences tailored to the specific needs of researchers, as we discuss below. The size of Science Bluesky, coupled with its decentralization and the extensibility of its underlying ATProtocol, make Bluesky/ATProto fertile soil in which to grow an ecosystem of transformative new open science tools.

Active knowledge creation over passive consumption

One area we are particularly excited to innovate in is the very concept of feeds. On one hand, we love the configurable feed ecosystem Bluesky is enabling, but at the same time, we, along with many others, feel we are running up against the limits of the medium. Feeds are inherently one-dimensional, ephemeral, and biased towards virality, creating attention dynamics where nuanced takes and valuable insights get buried beneath the posts with the most engagement. More importantly, they're optimized primarily for passive consumption rather than the active, non-linear and iterative process that knowledge building actually requires.

Collaborative tools for thought like Are.na point toward what becomes possible when we design for more flexible forms of knowledge creation and curation instead: researchers can make lists of related papers (like playlists on Spotify), share assessments (imagine Goodreads for science), and organize shared collections around research questions. These aren't just different interfaces - they're fundamentally different modes of eliciting knowledge and engaging with it that feeds are not designed to support.

The interoperable nature of ATProto makes this experimentation particularly exciting. Rather than choosing between feeds and curation tools, we can build systems where they reinforce each other. Semble can draw from the rich data already flowing through Bluesky while feeding structured knowledge back into the ecosystem to improve discovery and connections across apps.

Curation tools also offer a natural route to sustainability through paid subscriptions for premium features, as demonstrated by apps like Are.na and Sublime.

Growing interest in “micro-knowledge”, alongside stubborn adoption challenges

These new approaches to knowledge creation also connect directly to timely and important opportunities around nano/micropublishing. These concepts have of course been around for a while (we refer to "micro-knowledge" and nano/micropublishing interchangeably here), but they are recently seeing an exciting surge in interest. Progress in AI reasoning agents is driving the realization that models need access not just to “sanitized” papers, but also the messy thought processes by which we arrived at them. As Astera co-founder Seemay Chou recently wrote,

Scientists should probably be putting out shorter narratives, datasets, code, and models at a faster rate, with more visibility into their thinking, mistakes, and methods. In this age of the internet, almost anything could technically be a “publishable unit.” […] AI agents can’t learn scientific reasoning based on a journal record that presents only the glossy parts of the intellectual process, and not actual human reasoning. We are shooting ourselves in the foot.

That said, nanopublication is still quite niche and faces steep adoption challenges: there is still a critical lack of incentives for scientists to actually engage in such knowledge sharing. We think social media is a key piece of the incentive puzzle: as we’ve previously observed, researchers posting on social media are actually already nanopublishing but don’t know it yet. In other words, social media reveals the incentives that actually motivate researchers, for example engagement, reach, and meaningful interaction with peers. Framing also matters - as we discovered in our first experiment with the social media-to-nanopub autoposter, explicitly treating social media posts as "nanopublications" can actually deter researchers from participating (v1 in the diagram).

We think the key is finding the sweet spot combining the benefits of informal social sharing with the benefits of more formal publication, such as FAIR data. Or in other words, as depicted in the diagram below, the goal is to create apps that are delightful for both humans and machines (top right quadrant). Great UX is always a challenge, but there is a large space to explore here, full of promising yet-to-be-discovered tools. Importantly, ATProto provides an excellent testing ground for this experimentation, offering the technical flexibility to explore new experiences while building on the social networks researchers are already using.

We're aiming for the sweet spot combining the benefits of informal science social media sharing with the benefits of more formal publication

Bridges to Science Twitter

Where does all this leave Science Twitter? When we started the project, “way back” in early 2024, Twitter was the center of gravity for science social media, and we had been focused on “liberating” tweets from the platform by converting them to Nanopublications. A lot has changed in the last year, and Science Twitter has shrunk considerably. The user experience for those who remained has deteriorated considerably and the developer experience is even worse. Given the current platform turbulence, Science Twitter’s future is not clear.

With all that said, Science Twitter is currently still a significant community and it still harbors valuable discourse. We aren’t abandoning Science Twitter, but we are changing our priorities and approach to it.

One direction we’re particularly excited about is a collaboration with Community Archive, a wonderful project also focused on Twitter data liberation. Currently, users need to manually upload their data to the Archive, but tools to automate this process are also under development. An interesting potential collaboration interface could be an Archive - Cosmik - ATProto integration to enable public archiving while respecting user data sovereignty.

Movements and ecosystems over products

From the outset, we’ve recognized the need for ecosystems, beyond products. Since then, this recognition has only been sharpened by recent developments.

By the emergence of AI coding tools making product development more accessible than ever before.

By the increasing precarity of doing science in a fragmented, atomized and extractive system where researchers compete for dwindling resources while their labor enriches platforms and publishers that give little back to the scientific community.

By inspiring collaborations with other Open Science builders who share our conviction, like Rowan Cockett at the Continuous Science Foundation:

We started a movement because we need a place big enough to imagine something different—and small enough to actually build it together. Modular science isn’t just a technical shift, it’s a cultural one. And that starts with community.

By our collaboration with Ida Josefiina at The Agency, who helped us realize we’re building a movement. Products “can’t inspire people to feel like they are part of a cause they believe in. That’s what movements do.”

The real magic will happen when our systems help researchers stop feeling like isolated competitors and start experiencing themselves as part of a movement - a living, evolving collective intelligence. When unique insights don't disappear into the void but become building blocks for others, when our data trails trigger serendipitous connections between minds that would never have met - this is where breakthroughs will emerge.

Join the movement 🤝

]]> cosmik@newsletter.paragraph.com (Ronen Tamari) cosmik@newsletter.paragraph.com (Cosmik: Collective Sensemaking Networks) cosmik@newsletter.paragraph.com (Shahar Refael Oriel) cosmik@newsletter.paragraph.com (Wesley Finck) <![CDATA[Overcoming information overload with circular attention economies]]> https://archive.cosmik.network/circular-attention ZJqfgSuicHtCmiFVr0RA Fri, 30 May 2025 16:34:44 GMT Herbert Simon, in the oft-quoted passage from the 1971 piece “Designing Organizations For An Information-Rich World”, foresaw how information overload would transform attention into one of the most valuable resources:

In an information-rich world, the wealth of information means a dearth of something else: a scarcity of whatever it is that information consumes. What information consumes is rather obvious: it consumes the attention of its recipients. Hence a wealth of information creates a poverty of attention and a need to allocate that attention efficiently among the overabundance of information sources that might consume it.

Over 50 years have passed and internet scale data has made this problem more acute than Simon could have ever imagined. Yet our research infrastructure does not reflect this reality. What research merits attention? What papers, trends and discussions should I, as a researcher, be paying attention to? For all the awareness of the problem, there is still a surprising lack of tools helping researchers answer these pressing questions. For most researchers, information overload results in a constant feeling of falling behind the exponential growth of knowledge in their field, too many open browser tabs and not enough ways to decide which ones to close and which to read.

Traditional curation methods can’t keep up…

Journals used to be the answer; traditionally, they were tasked with filtering out poor research and curating high quality work. However, journals have fundamentally failed to scale with the explosion of research output, often taking years to review and publish while knowledge evolves at internet speed. Moreover, commercial incentive structures have warped their filtering function: “what science gets done and how it is presented, has been distorted by what is profitable for publishers”.

… AI alone isn't enough either

Can’t we just ask Claude or ChatGPT? Despite remarkable recent advances, AI alone is not a solution either. To see why not, try to imagine Goodreads or IMDB reviews where all the reviews are AI generated. Though Mark Zuckerberg might not see any problem with that, for most of us, something would be missing. AI will certainly be part of the solution, synthesizing, aggregating and organizing information, but there is a fundamentally social aspect of attention that cannot and should not be automated. Shared attention is part of what makes us human. We care about what human peers and experts find interesting, not just what opaque algorithms recommend.

The promise and limits of social media

The social nature of attention is also attested to by the widespread use of social media by researchers. Ask a researcher how they find interesting papers to read, and more often than not, social media figures largely in their answer. And yet, social media is not designed for science; research content is buried in noisy feeds, platforms optimize for engagement and not sensemaking, data is siloed and fragmented, and the medium is not expressive enough for complex knowledge work. Despite its limitations, social media provides an important clue towards how we can overcome information overload.

From linear to circular: rethinking the attention economy

To better understand the role of social media in guiding individual and collective attention, let's revisit Herbert Simon's statement of the information overload problem. In his account, the attention economy is linear: attention is a resource consumed by information, lost forever - like fossil fuels burned to generate energy. But what if we thought about attention differently? In contrast to a linear economy, a circular economy aims to create systems where resources are continually reused and recycled, minimizing waste and maximizing efficient resource utilization. The circular economy framing suggests a follow-up question to Simon: information consumes attention, but what does the consumption of information produce? In other words, what are the resulting artifacts from the process of attending to information?

The untapped creative exhaust of research

Classic examples of attention artifacts are notes or annotations in personal knowledge tools like Obsidian, Notion or Zotero. Sometimes there may be no material "attention traces" - perhaps a paper wasn't worth finishing or taking notes on. And obviously, in some cases insights will be kept private. But more and more researchers are sharing insights publicly, for example as posts on social media recommending or critiquing something they recently read.

These short shared insights - public products of human attention - are like ant pheromone trails, crucial signals for shaping collective attention. They represent a more circular attention economy at work, where one person's consumed attention becomes a signal that guides others' attention allocation.

But such sharing is still few and far between. How much of what we read do we actually share publicly, even when privacy is not an issue? How much do we share in open and interoperable formats, vs sharing to platform siloes? Currently, most of our attention is consumed and not recycled into useful signals for the broader research community. Joel Chan, a researcher at the University of Maryland, talks about the "untapped creative exhaust" of millions of researchers already pouring over hundreds of millions of references - a vast repository of human sensemaking that remains largely invisible and un-utilized. A.J. Boston, a scholarly communications librarian at Murray State University, has also written extensively about the great (yet unfulfilled) promise of harnessing internet-scale energy for open micro-reviews: “Considering the web’s tendency for moderation and review, it’s honestly a little weird we don’t already have established practices in place to more systematically enable, capture, and display open comment on research objects”.

Co-augmented reality

Imagine if digital attention sharing was more of the norm. The fellow in the meme would have access to a kind of “co-augmented reality glasses” to overlay their bookshelf, surfacing previously hidden trails of knowledge and insight shared by others:

The Goodreads mobile app demonstrates this kind of co-augmented reality at work. You point your phone at a book and pull up human reviews about it. This works impressively well, both in terms of the coverage and the quality of reviews. A kind of Goodreads for research, though it might differ in implementation details (e.g., a five star rating system may be overly simplistic), would clearly be incredibly valuable. AlphaXiv is an example of some great work being done in this direction of Science Goodreads, specifically for arXiv preprints.

Building infrastructure for circular attention economies

The opportunity is clear: we need systems that can represent, aggregate, and redistribute the traces of human attention in ways that help the entire research community navigate information overload more effectively. This means building tools that make it easy and worthwhile for researchers to share their attention traces - whether that's a short review, an annotation, or a recommendation - and then intelligently routing those signals to others who would benefit from them.

Such a system would create positive feedback loops: the more researchers participate by sharing their attention traces, the better the system becomes at helping everyone find relevant, high-quality information. Social feedback would allow participants to see their contributions making a tangible difference, thus encouraging further sharing. Building this infrastructure according to FAIR (Findable, Accessible, Interoperable, Reusable) data principles also means that, rather than be captured in platform siloes, attention data can easily flow across the ecosystem to inform other content recommendation and discovery services.

It's a fundamentally different approach from both the traditional gatekeeping of journals as well as pure algorithmic recommendation - one that harnesses the collective intelligence of the research community while respecting the profoundly human aspects of attention and judgment.

There are still many open questions to address as we build this infrastructure:

Incentivizing participation: What tools and social mechanisms can make sharing attention traces feel rewarding rather than burdensome? How do we move beyond the current reality where most insights remain private?

Quality and moderation: How do we balance open, permissionless sharing with the need to maintain signal quality? What approaches can prevent gaming or manipulation while preserving the authentic human judgment that makes these traces valuable?

Interface design: Reviews and curation come in many forms—faceted reviews, star ratings, annotations, recommendations. What are the right schemas and interfaces that balance expressiveness with ease of use, avoiding both oversimplification and overwhelming complexity?

Aggregation and discovery: How do we synthesize thousands of individual attention traces into coherent signals that support effective navigation without creating new forms of information overload?

Data sovereignty and ownership: How can we preserve users' data sovereignty and provide an alternative to the extractive practices that dominate today's platforms, where user-generated content is harvested without meaningful consent or compensation? What models ensure that researchers retain control over their attention traces while still enabling collective benefit?

Navigating the perils and promise of the digital age requires, according to author James Williams, that “we give the right sort of attention to the right sort of things... A major function, if not the primary purpose, of information technology should be to advance this end.”

Building circular attention economies represents one promising path toward that goal - turning our attention from a scarce resource into a renewable one that compounds in value as more people participate.

We’re excited to be attending to attention and pursuing these questions at Cosmik. We also think the best way to address them is together, in community. In the next few months we will be exploring new ways to foster these discussions. Meanwhile, feel free to drop us a line or find us on social media!

]]> cosmik@newsletter.paragraph.com (Ronen Tamari) cosmik@newsletter.paragraph.com (Shahar Refael Oriel) <![CDATA[From SenseNets to Cosmik]]> https://archive.cosmik.network/cosmik-intro ghQpYpiZqY3rSFfnVOuU Thu, 20 Mar 2025 14:01:34 GMT Dear sensemakers,

The last few months sure went by in a blur. In March, we officially wrapped up the Astera Open Science fellowship. The ending of a good period is always bittersweet, but at least now we have time to unpack and share some learnings and updates. This post will provide a broad outline, and in future posts we will go more in depth on some of the topics raised here.

SenseNets → Cosmik

First, we changed our name to Cosmik! We realized SenseNets happens to be the name of a Chinese company developing facial surveillance technology, so we decided to rebrand. Cosmik stands for Collective Sensemaking Networks, and also reflects our ecological stance towards science and intelligence more broadly. Kosmos is "world" or "universe" in Greek. Intelligence is not just in our heads, but also deeply rooted in our interactions with the physical and social worlds around us. Our thinking mediums are the message. Environments can degrade cognition or they can make us smarter, and we’re on a mission to design and create environments that augment human capacities for collective sensemaking. Our new home on the web is cosmik.network, and you can also find us on Bluesky, Mastodon and Twitter. Check out our new 2 minute intro video presenting Cosmik!

v0.2 launch: Hyperfeed

At the end of 2024 we launched a new alpha version of the Cosmik app, v0.2, to a few dozen researchers. v0.2 featured the Hyperfeed, a new kind of AI-enhanced, cross-platform research feed. The previous version of the app, v0.1, was geared towards helping researchers convert their social media posts into nanopublications. However, as described in the last blog post, through user testing we learned that more work is needed to establish tooling and incentives for nanopublishing.

With the Hyperfeed, we pivoted from publishing to focus more on the information discovery experience. Browsing a feed requires less effort from researchers and provides more immediate value. Underneath the hood we continued to use our LLM pipeline to convert posts to the the structured knowledge graph nanopublication format, and it was exciting to experiment with user experiences leveraging these semantically enriched posts.

There is still work to improve the UX, and we are missing some key features which will be necessary for an open beta for Cosmik, in particular feed personalization based on a user’s social graph.

That said, we see great potential in feeds powered by rich knowledge graphs. One particularly cool knowledge-graph enabled feature is a “semantic layer” over references: for any reference shared to the network, the Hyperfeed aggregates all of the posts mentioning that reference, along with their semantics (which posts agree with the reference, which disagree with it, and so on). In this way, users get a sense of the general sentiment around a reference at a glance, and they can also easily navigate the diverse perspectives relating to it. The figure below shows the how Cosmik’s semantic layer (right) provides additional context missing in traditional social media feeds (left).

Cosmik’s semantic layer (right) provides additional context missing in traditional social media feeds (left). Source post: https://bsky.app/profile/carlbergstrom.com/post/3ljorhabhkc23

Open science sustainability

We have also been busy thinking about sustainability models for Cosmik. Creating a successful startup is already hard enough, and we learned that creating an impactful open science venture poses unique challenges on top of that. Open science is a public good. Correspondingly, infrastructure supporting open science should uphold additional standards around governance and sustainability, as well as insurance against data loss or enclosure (the POSI Principles). Some of these standards, like open data or open source code, can be at odds with the business incentives of startups who must survive in highly competitive environments. Venture-backed startups also face particularly challenging decisions about exit strategies, as the conventional path of acquisition by publishers may satisfy investors, but frequently undermines open science principles.

The non-profit route is not a panacea, either. Non-profits generally depend on philanthropic donations which require extensive application processes, and rarely provide the stable, long-term financial support necessary for sustained operations. Recent political shifts in the United States demonstrate the vulnerability of even government-funded scientific infrastructure.

For Cosmik, we identified a novel organizational and sustainability model in the form of researcher-governed data co-operatives. This model is inspired by Subvert, a similar effort in the music industry, and also the insight that the type of data shared by scientists on social media can be extremely valuable when aggregated and integrated with analytics services, both to other researchers as well as to institutions.

Co-operatives, while not common (yet!) in the landscape of open science ventures, elegantly address many key challenges by combining the sustainability features of for-profit entities while maintaining stakeholder alignment:

  • Researchers have a stake in the value they generate, incentivizing their participation

  • Community governance and ownership solves the exit problem and provides resilience against predatory acquisition

For Cosmik, we are pursuing a strategy of philanthropic and other POSI-aligned funding, with a long term goal of creating a self-sustaining eco-system with revenue from membership fees, data-sharing agreements, and value-added services such as advanced AI features.

More on the data co-op model in future posts!

A Cosmik moment for researcher-governed science social media

A lot has happened in the fellowship year, including our own learnings as well as external world events. Taken together, we are at a unique moment now where a lot of stars are aligning for our vision of Cosmik and the idea of a new kind of social network for researchers.

Researchers want better social curation & discovery tools

Posts about “goodreads for science” have a curious tendency to go viral. Clearly, it’s on a lot of people’s minds. Check out this post from just a few months ago, which triggered huge discussions:

We also heard similar sentiments firsthand in dozens of interviews with researchers: for many, social media is the place to stay abreast of new trends in their field, but they also expressed frustration with noisy feeds and fragmented discussions.

Researchers want to publish insights, not just papers

We recently came across a Nature paper citing Twitter posts, which encapsulates both the brilliance and the brokenness of science social media. On one hand, insights shared on social media are found to be worthy of citation in a Nature paper, no less! But at the same time, we can and should do a lot better than closed platforms like Twitter in terms of a publication format for small units of scientific knowledge (Nanopublications, anyone?).

More broadly, during the fellowship we learned that many researchers, especially those outside of traditional academia, want to participate in science, but they don’t have the time or incentives to write full papers. These researchers have valuable insights to share, but they lack the right medium for their contributions. As a result, we are losing vast amounts of potential insights to “creative exhaust”.

Researchers increasingly care about infrastructure, governance and data rights

The extractive nature of the academic publishing ecosystem is not news for most researchers, who have nonetheless accepted these conditions as an unavoidable cost for doing science.

Efforts to try and change the “rules of the game” happened mostly on the margins. However, recent developments - such as the decimation of Science Twitter following Musk’s takeover, publishers brazenly selling researchers’ labor as training data to GenAI companies, and the Trump administration's cuts to federal science funding - are jarring the status quo. The silver lining to all this chaos is that it is creating space for new approaches to be tried, and it is also catalyzing discussions around how to develop sustainable, resilient, and community-governed research infrastructure.

Massive growth in open science social media (Bluesky)

One very real change in the status quo is the exodus of scientists from Twitter to Bluesky. Altmetric has reported that on some days the volume of research mentions on Bluesky now equals that of Twitter, despite Bluesky being at least 10x smaller. The formation of Science Bluesky presents a unique opportunity: due to its extensible and decentralized nature, Bluesky and its underlying ATProto provide fertile soil for an ecosystem of transformative new open science tools. For example, we are already seeing ATProto Goodreads apps cropping up, and an ATProto Science Goodreads is a natural evolution of this trend!

End of a chapter, start of a journey

The fellowship has ended, but in many ways the journey is just beginning.

We’re continuing work towards an open beta for Cosmik: we’re developing a deeper integration with Bluesky and ATProto, and we’re also exploring ways to improve the discovery experience with social graph based personalization and semantic search.

On the organizational side we are continuing to develop the researcher co-op and are aiming to open Cosmik for membership when we finish incorporation.

To support these next steps, we are actively fundraising from philanthropies and POSI-aligned investors.

The Astera program was an amazing opportunity to take on big questions at the bleeding edge of science, how we publish it, how we communicate it and how we make sense of it. As far as I know, there is no other program like it in terms of the generous support and amount of independence provided.

I also recognize that big aspirations come with the risk of falling flat. At times I feel a sense of vulnerability that the tangible outcomes of the fellowship may seem modest - we haven’t even launched an open beta yet. And yet, we're playing the long game. While we recognize the crucial role of better technology, we also know that tech alone won’t save us. We believe that truly transforming how we do research will go beyond specific apps and tools, and that our efforts must tap into researchers' deeper aspirations for a more equitable, participatory and empowering knowledge ecosystem. Building this kind of researcher-governed infrastructure requires patience and persistence - it means questioning established systems and working with change leaders to propose and implement bold alternatives. The path is challenging, but the potential to transform how science functions makes the journey worthwhile.

Big thanks to the Cosmik team: Pepo, Shahar, Wes, Andrea and Colin. And special thanks to Astera and the Open Science team for making all this possible, in particular Jessica Polka, our program director, and Prachee Avasthi, head of Open Science.

]]> cosmik@newsletter.paragraph.com (Ronen Tamari) <![CDATA[SenseNets alpha launch retrospective]]> https://archive.cosmik.network/alpha-launch lW9JT4jjOWxj0bdSW00L Wed, 06 Nov 2024 18:27:35 GMT In September we wrapped up a focused release of v0.1 of the SenseNets app to around a dozen alpha testers. Our intrepid testers were mainly academic researchers who were already active on science social media and interested in open science and novel publication methods.

SenseNets v0.1 release: AI nanopublishing assistant

This release was centered around the experience of converting social media posts to nanopublications: users could connect their Twitter accounts, and we processed their posts with AI to (1) flag potential nanopublications and (2) add metadata to make the posts easier to find. Users could then review and edit the drafts, and finally nanopublish the posts they wanted to. We also enabled automated nanopublishing for users that preferred to skip manual review.

You can see a short demo of this version here:

What makes a social media post nanopublication-worthy?

One of the key questions that comes up around nanopublishing is: “what makes a science social media post a potential nanopublication?” Since nanopublishing is such a new practice, there actually aren’t really many established norms yet. As detailed by the Nanopublications webpage, “a nanopublication can be about anything, for example a relation between a gene and a disease or an opinion.” We similarly adopted an expansive view of nanopublishing, treating any posts even remotely research-related as valid potential nanopublications. Nanopublications are part of the Semantic Web, of which one of the core pillars is the “AAA principle”: Anyone can say Anything about Any topic.

In practice, for automated classification of posts as potential nanopublications we decided to use two indicators, where the presence of either means we classify the post as a potential nanopublication.

  1. Reference-type: checks if the post contains any mentions of academic references (e.g., references that have a DOI or other scholarly identifier, like preprints or journal papers).

  2. LLM-based: classifies the post based on content. This indicator is less precise than (1) but helps detect a wider variety of research related content. For example, this indicator can detect a post announcing a informal blog discussing a new machine learning method.

In the next section, we'll discuss our findings and how our experiment played out in practice. For more technical details about our nanopublishing pipeline, check out the appendix section at the end.

What did we learn?

Researchers are curious about nanopublishing

While most of our alpha testers hadn't heard of nanopublishing, we observed a lot of curiousity about the idea. Researchers feel that a lot of valuable research-related knowledge goes unrecorded or unrecognized, and that this knowledge often takes the form of short claims, observations or assertions. We encountered similar curiosity about nanopublishing more broadly, also outside our focused alpha testers group. As just one example, a chance meeting with an astrophysicist in a meetup in Boston resulted in me being invited to present SenseNets and nanopublications at the Harvard Center for Astrophysics - they know social media is relevant for them as scientists but are also curious about how it could be improved through new practices.

Researchers still aren't sure what to nanopublish

That said, the novelty of nanopublications is also a source of confusion. We found that nanopublishing means a lot of different things to different people. Our working definition of nanopublications has been any research-related social media post, as described above. But some of our testers took a stricter view, and saw nanopublications as more like micropublications - an intentional kind of microblogging around research. It was not obvious to testers that merely the statement that one has read a particular paper could be a nanopublication. A recurring theme was that nanopublishing involved more friction than merely posting about research on social media. One reason for this impression is simply the name (publishing vs posting) and another reason is that nanopublications behave more like publications in the sense that they can only be retracted, not completely deleted (like a post can).

Better incentives are needed for nanopublishing

Nanopublications' novelty also means there still aren't enough incentives for nanopublishing. We hoped that support for Open Science practices and "data altruism" would prove motivating enough for these early stages, but our testers made it clear that nanopublishing has to be more useful for them to invest time in doing it. For example, being able to view how one's nanopublications are being received, or being able to use nanopublications for new kinds of discovery experiences. One idea for a discovery service that a lot of researchers were particularly excited about is a "science only" feed.

Nanopublications inspire researchers to think of new ways of sharing information

One of the reasons we continue to be excited by nanopublishing is their capacity to inspire researchers to think of new ways to share their knowledge and connect with their peers. Multiple researchers reported that they started seeing their social media posts "with fresh eyes", as nanopublications. One researcher told us they are seeing how nanopublications can transform science social media into the leading edge of academic work in a space.

Some researchers had ideas around using nanopublications beyond science social media, for example to record citations in conference presentations.

Another surprising related finding was that some researchers were excited by the idea of using nanopublications instead of science social media. These researchers did not like to participate in social media, but felt like they had valuable knowledge to share with their research communities. They saw nanopublications as a new way to convey this knowledge.

Next steps

To summarize, this first experiment was focused on making nanopublishing easy, through the use of an AI nanopublishing assistant. Though limited in scope, the v0.1 release already taught us a lot about the current limitations and promise of nanopublishing. Drawing on our learnings, our next experiment will focus on further reducing the friction around nanopublishing, and also making nanopublications more useful and engaging. We came away with some exciting ideas for how to do that, and aim to release an updated version of the app in the coming weeks - stay tuned!

Special thanks to our designer Andrea Farias for user research and the demo video, and thanks again to our alpha testers for providing invaluable insights and feedback!

Appendix: unpacking the nanopublishing pipeline

The diagram below describes our nanopublishing pipeline:

  • The platform watcher service automatically fetches new posts as the user creates them.

  • Posts that are part of a longer thread are merged, and mentioned URLs (references) are extracted and normalized.

  • We use the Citoid API to extract reference metadata, including the type of content (blog post, journal article, etc) and citation information like DOIs.

  • Posts are then processed with our NLP module which is composed of multiple Langchain modules.

    • A research detection module flags posts as potential nanopublications using a combination of reference metadata and AI, as described above.

    • The keywords module extracts keywords related to the content and references.

    • The relation tagger labels the semantic relations between the posts and the references it mentions using our Common SenseMaking Ontology (CoSMO). CoSMO draws on a number of related ontologies including the Citation Typing Ontology CiTO and Schema.org. CoSMO also includes new concepts specific to science social media, such as asksQuestionAbout or summarizes.

  • To nanopublish, we used the nanopub-js library

]]> cosmik@newsletter.paragraph.com (Ronen Tamari) <![CDATA[From Genius Science to Scenius Science]]> https://archive.cosmik.network/scenius-science lX9AsUsVhmTgWqm0UFtZ Wed, 31 Jul 2024 00:00:00 GMT (cover image from Matt Akamatsu's excellent talk at DeSci Denver 2024)

Hello sensemakers!

We’ve been hard at work over the past few months preparing the first alpha release of Sensemaking Networks. We’ll share more updates about our progress soon, but this will kick off a series of posts where we will share some of our learnings as we’re engaging with the decentralized/open science research communities around the project.

For those who haven’t seen the previous episodes, Sensemaking Networks (or SenseNets) is a collective intelligence system designed to radically enhance information sharing and collaborative knowledge synthesis among researchers by integrating decentralized semantic web technology with AI and social networks. In our work on SenseNets over the past few months, we have come to notice a puzzle at the heart of how we do science.

The puzzle, which we will explore in this post, is that many scientists value collaboration with other researchers, but the tools and metrics we use isolate us and promote competition. Why is that?

To feel into this tension, imagine you are doing a literature survey of a new field, and have a stack of research papers in front of you. You have to decide which ones to read and in what order. To help you decide, you have access to assessments of each paper, but you need to choose between GPT-generated assessments or human assessments made by peers and experts in their field. Which would you choose?

I asked the audience this question during a presentation of Sensemaking Networks at an Open Science night organized by Astera in May. I was surprised by how few people preferred the GPT option (hey, this is the Bay Area, after all 🙃).

These two choices (GPT and peer assessments) represent two different approaches to sensemaking in science: the former is focused on empowering individual researchers in isolation, while the latter focuses on empowering collective, networked and social aspects of sensemaking.

Sure, phrasing the question as a binary (either one or the other) is a little unfair. Indeed, we are most excited by synergistic combinations of both approaches. However, the story is a useful demonstration of the puzzle we opened with: people actually really value social sensemaking, and yet technology is overwhelmingly tailored to individual sensemaking. Personal AI research assistants are a dime a dozen, while tools that help researchers share information and think better together are few and far between. The very lack of recognition for the role of social media in science is the raison d'être of Sensemaking Networks.

In fact, this isn’t just the tech focus: academic research itself is also primarily tailored towards augmenting individual researchers. Recent research even proposed developing a personalized “science exocortex” using a swarm of AI agents to automate an ever-growing range of scientific activities while leaving just the highest level decisions for the human operator.

What’s gives? Why are we so focused on the individual side of science? And, given that we intuitively really value working with other humans, why don’t we have more tools that improve collective sensemaking?

It turns out that there are a lot of intriguing answers to these questions. We’ll take a whirlwind tour through sociology, meta-science and collective intelligence theory, and reflect on what these diverse strands of research might imply for the future of SenseNets and science information systems more broadly.

Sociology and lone genius mythology

Part of the answer lies in the “lone genius mythology” around science. We were all raised on stories of genius scientists who defied their doubters and against all odds made discoveries that advanced the frontier of human knowledge. These stories weren’t wrong, of course; these scientists were extraordinarily talented and determined individuals. Yet the stories are also incomplete, for example omitting reference to other crucial (e.g., social) factors that contributed to the process of discovery. An insightful critique of the recent Hollywood portrayal of Robert Oppenheimer contrasts Hollywood’s “lone genius and his blackboard science” with a perspective inspired by French sociologist Bruno Latour, according to whom “’The father of the atomic bomb’ was no single parent, but rather a collective, networked one…“

In other words, the developers of sensemaking tools for individual researchers might be drawing more inspiration from Iron Man than from Latour.

But the focus on individual researchers is not just a product of the stories we are told about scientists; there are more pervasive forces at play.

Meta-science: Re-thinking the pecking order in science

A thought-provoking recent paper called “Shifting the Level of Selection in Science” shows how an overfocus on individuals is built into the very fabric of science, by way of the reward structures used to evaluate scientists:

The predominant approach to scientific evaluation uses individual-level criteria, such as one’s number of first-authored publications, citations, h-indices, journal impact factors, and success in funding acquisition… This evaluation strategy implicitly assumes that identifying and rewarding the most accomplished individuals is the best way to generate scientific knowledge.

The authors show how individual-level criteria encourage competition and personally beneficial behavior among scientists. But, as they observe,

personally beneficial behaviors are only a subset of the behaviors that benefit science. For example, the scientific community plausibly benefits from the open sharing of information such as code, materials, and raw data, whereas individual-level competition disincentivizes information-sharing to hinder competitors’ success

It turns out that we have a lot to learn from chicken farmers:

the most productive hens in a coop are also the nastiest hens, feather-pecking and cannibalizing the other hens in their coop. Because individual hens who are most productive are those that harm others, selectively breeding the most productive hens can actually lead to lower overall egg production

Instead of selecting the most productive individuals, breeders have learned to select the most productive groups.

Perhaps science can do the same? The authors make a solid case for expanding evaluation mechanisms to incentivize group-level outcomes in addition to individual-level metrics. Importantly, such evaluation will require far greater recognition of diverse prosocial contributions, such as information sharing and other currently invisible “team science” roles.

So we have another piece of the puzzle - in a world where researchers are being incentivized to conceal rather than share their insights, private AI assistants are a safer bet for tool developers, even though social sharing platforms might actually be more effective in advancing science.

Science as process, not product

Another fascinating line of research goes even more existential, in questioning the very aim of science. “Shifting the Level of Selection in Science” proposed new group-level incentives, but still focuses on a productivity-based model where science is about producing new knowledge. A recent piece called “An Epistemology for Democratic Citizen Science” makes a compelling case that sometimes the goal is the journey in science. Rather than just viewing the knowledge as the product of science (what they call “industrial science”), they make the case for re-thinking the social and cognitive processes that are generating scientific knowledge (”ecological science”). Ecological science, beyond traditional science’s role as an inquiry into the natural world, is also “an inquiry into how to best cultivate and utilise humanity’s collective intelligence”.

On this view, the social tools we use to communicate about science would themselves be at the core of the scientific process. Those tools would function as large scale experiments in collective intelligence (CI), informed by the latest research. Science Twitter is cool, but leaves much to be desired as a tool for researchers. We have a lot more to learn from CI theory; in “Science Communication as a Collective Intelligence Endeavor”, the authors provide an outline of what CI systems for science might look like. In particular, such systems would (a) enable better aggregation of distributed knowledge, (b) involve a more diverse group of contributors and (c) encourage increased public participation in science.

Connecting this research to our opening question, perhaps many tool developers are building “industrial science” tools for enhancing individual productivity, as opposed to “ecological science” tools for enhancing collective intelligence.

Industrial science, like industrial agriculture, risks creating scientific monocultures where

some types of methods, questions and viewpoints come to dominate alternative approaches, making science less innovative and more vulnerable to errors.

In contrast, ecological science is more like permaculture, reflecting a diversity of methods, contributions and perspectives. See Matt Akamatsu’s talk at DeSci Denver for an exciting example of what this might look like

Conclusion: from Genius Science to Scenius Science

For those of you who haven’t heard of the term “scenius”, musician-activist Brian Eno coined it some 30 years ago:

I became (and still am) more and more convinced that the important changes in cultural history were actually the product of very large numbers of people and circumstances conspiring to make something new. I call this ‘scenius’ - it means ‘the intelligence and intuition of a whole cultural scene’. It is the communal form of the concept of genius.

Curiously, scenius has been around for a while but actually hasn’t taken off:

But given the dominant focus on individuals that we’ve seen, maybe we shouldn’t be too surprised at this point 🤷

Nonetheless, it feels like genius is fading and scenius is in the kairos. Converging evidence from across multiple fields suggests that if we’re really serious about enhancing science for the benefit of humanity, we should be thinking less about how to find geniuses and a lot more about how to create sceniuses.

This post was inspired by the Sensemaking Scenius, our own scenius in the making 🏗

]]> cosmik@newsletter.paragraph.com (Ronen Tamari) <![CDATA[Sensemaking Networks Part 3: building partnerships and prototypes]]> https://archive.cosmik.network/pt3-prototyping 22cdRi50U0DXSBb2HFvk Mon, 11 Mar 2024 00:00:00 GMT Hello again fellow science sensemakers!

We just got back from DeSci Denver and have a lot to unpack, both from the conference as well as from the last couple of months. To TLDR - interviews with researchers inspired us to develop a semantic cross-poster that can nanopublish alongside posting to social media. We built our first prototype and presented it at DeSci Denver. We also created an LLM-augmented interface for querying the nanopublications network, with an eye towards enabling next generation natural language knowledge interfaces. “Sensebot, summarize the debate around this paper, and show me the positions of the researchers in my network” (ok, we still don’t support that query yet, but we’re working on it :) )

Sensemaking Networks crew all together IRL for the first time, in beautiful Colorado! From left: Pepo, Ronen and Shahar

Product Design & User Sensemaker Research

Since January we’ve been meeting researchers from all walks of science to learn how they use social media for their research, and how they’d like to improve it (we prefer to call them (sense)makers, not users!)

These sessions helped us appreciate the great diversity of ways in which researchers acquire and share information. For example, some avoid social media in favor of personal knowledge tools like Roam Research, while others use their social media account itself as a public knowledge repository of sorts. Some researchers use cross-posters as a way to deal with social media fragmentation; cross-posters publish their content simultaneously to multiple platforms.

We also learned that open social media does not necessarily mean an invitation to scrape data: in some science Mastodon communities for example, bots are actively discouraged, even for research purposes.

From scraping to self-sovereign semantic cross-posting

These learnings informed our design process and led us to update our project plan. Rather than building another social media crawler, we decided to start with a semantic cross-poster; a cross-poster that enables authors to easily publish to the Nanopublications network, adding rich machine-readable semantics to their posts, beyond just text and media. A semantic cross-poster is relatively simple technically, making it an attractive starting point. Furthermore, it aligns with the human-centered design principles guiding our project:

  • Cross-posters preserve consent: people are subjects that author their posts rather than data point objects to be scraped unknowingly, as in services like Altmetric.

  • Cross-posters “meet people where they are” by integrating with the tools and practices they already use. Beyond social media integrations, cross-posters can also interface with personal knowledge systems.

  • Semantic cross-posters enhance people’s agency: semantics increase posts’ discoverability by making them FAIR (Findable, Accessible, Interoperable, and Reusable) and provides authors more control over how the algorithm “sees” them. Instead of outsourcing decisions to opaque AI algorithms far off “in the cloud”, authors are participants in decision-making loops and control how their data is intended to be read by machines. FAIR and semantic data also enhance collective intelligence by connecting posts in a distributed knowledge graph that can power customizable feeds and recommendation systems.

  • Cross-posters save researchers time and energy by enabling them the option of automating potentially tedious tasks like adding semantics, making data FAIR, and manually posting to different platforms. A planned feature we’re particularly excited by is auto-posting, enabling authors to define a single input source that will automatically be monitored and cross-posted from whenever new content is published on it. For example, authors will be able to automatically cross-post from their Twitter account whenever a new post about research is detected (with the option of a quick manual vetting to correct errors).

Semantic auto-cross-posting effectively enables a kind of “self-sovereign scraping”. Traditional scraping happens in the background with no user involvement. While the fact that it requires no user attention may be seen as convenient, it also seriously limits their control. There is no easy way to opt out of scraping, and also no way to influence what data is being collected about you. In contrast, our semantic cross-poster will enable people to benefit both from convenience and also control over what data they share, and how.

Prototyping

Screenshot from our first semantic cross-poster prototype!

In February we built our first prototype, which supports cross-posting to Twitter and the Nanopublications network. Semantics are extracted using a Langchain LLM pipeline (we experimented with the Mistral open source model and GPT-4). We are also experimenting with LLM pipelines for automated keyword extraction. For nanopublishing we use the nanopub-rs library(https://github.com/vemonet/nanopub-rs), and we streamlined the private key management using web3 wallet services. The screenshot to the right demonstrates an example, and you can check out this video for a quick walkthrough (see the resulting nanopub here). We’ll share more technical details as we stabilize the app.

Making sense of sensemaking

Representing Semantic Web concepts requires an ontology, which describes the concepts and possible relations between them. Our ontology is called Common SenseMaking Ontology (CoSMO). CoSMO is designed to represent the diverse ways in which researchers relate to research on social media. CoSMO is based on the classic Citation Typing Ontology (CiTO) which characterizes citation relations in scholarly works. For example, CiTO includes concepts like agreesWith, repliesTo, and citesAsRecommendedReading. Where CiTO is focused on more traditional scholarly citations, CoSMO is an updated adaption for sensemaking on digital social networks, including concepts from CiTO but also support for other kinds of media like podcasts and videos, as well as mentions of events or job opportunities. The current version is just a start, and we expect to evolve CoSMO as we deploy the cross-poster in actual communities. In the future, we’d also like to enable communities to easily extend and adapt ontologies to their specific needs.

Querying nanopubs using LLMs and SPARQL

As discussed in previous posts, nanopubs implement the FAIR standard. A key to nanopubs’ Findability is the SPARQL query language, which provides a powerful tool for querying the nanopublications network. Queries can power diverse range of applications such as configurable feeds, recommenders and question answering systems. Query languages such as SPARQL have existed for years but remain underutilized because they're technically demanding. Like with semantic publishing, we're enthusiastic about the potential of using large language models (LLMs) to simplify access and create powerful yet user-friendly, natural language-driven interfaces.

From some preliminary experiments we can already see great potential. For now, we’ve tried simple queries like “show me all critiques of this paper” and “show me all paper recommendations from this list of researchers”.

We’re excited by the potential of more complex queries powering next-generation knowledge interfaces. Interfaces that used to be firmly in the domain of sci-fi are now an imminent possibility:

“summarize the debate around this paper, and show me the positions of the researchers in my network”

“show me the people in this conference who have read any of my papers and enjoyed them”

“keep me notified on all job offers in biodiversity research in institutions which are affiliated with any researchers in my network”

Interestingly, these kinds of queries are simply not possible even with the best LLMs today, since they rely on structured semantic knowledge graphs in addition to powerful NLP capabilities. “chat with paper” bots are cool, but we think that combining LLMs with nanopubs unlocks a whole new class of interfaces grounded by semantics and attributable to actual human comments. Such systems will be much less prone to hallucination, and will help connect between human researchers rather than disintermediating them through unreliable and opaque models.

Sensemaking Networks @ DeSci Denver 2024

Our whole team met in DeSci Denver at the end of February to present and demo Sensemaking Networks (slides). It was an exhilarating week, and our heads are still spinning from all of the new connections and ideas that emerged from the conference!

Ronen presenting Sensemaking Networks in the DeSci event at Colorado University in Denver

We were especially energized by discussions around coordinated efforts to bring DeSci and open science projects more into the mainstream. Imagine a movement of scientists opting in to collectively FAIRify their social media communications around science! Given the existing fragmentation across platforms, the situation can seem pretty hopeless at times, but we were excited to learn that we share these dreams with many others. John Lennon once said: “A dream you dream together is reality.” Shared dreams + great shared open science infrastructure=✨

If you’re excited by these dreams as well, we’d love to hear from you! Connect with us here if you’re interested in joining our beta when it’s ready, or any other collaboration!

]]> cosmik@newsletter.paragraph.com (Ronen Tamari) cosmik@newsletter.paragraph.com (Shahar Refael Oriel) cosmik@newsletter.paragraph.com (Pepo) <![CDATA[Part 2: Sensemaking Networks Project Plan]]> https://archive.cosmik.network/2-project-plan 8hg7Rax3Yc3eKijiTCtD Tue, 26 Dec 2023 00:00:00 GMT In the previous post, we surveyed the landscape of existing science communication infrastructure and how it motivates Sensemaking Networks, an integration of semantic nanopublishing and AI-augmented social media networks. Nanopublishing brings best practices of open scholarly infrastructure and leverages semantics to radically enhance interoperability and information flow. Social media apps bring usability, rapid feedback loops, and the power of algorithmic reach.

In this post, we'll begin discussing actual implementation plans.

🚧This project has a lot of moving parts, so we expect ongoing iteration to improve the plan! 🔄

Project Plan

We're currently planning two mutually supporting efforts: (1) recovering existing nanopublications from social media, and (2) building a social app for semantic nanopublishing based on open social networking protocols like ActivityPub and AT Protocol (the protocols underlying Mastodon and Bluesky).

1. Collecting nanopublications from science social media

As discussed in the last post, researchers are already nanopublishing on social media, but because the infrastructure doesn’t actually support it yet, valuable scientific knowledge is being lost through data enclosure and fragmentation. Fortunately, powerful NLP (Natural Language Processing, a.k.a large language models and friends) tools can help us recover some of this knowledge and make it available on the open scientific record.

This effort will focus on detecting mentions of research in social media posts, linking them back to their authors and inferring their semantics where possible. Recovering semantics from natural language is called parsing in NLP.

For example, consider this tweet on social media:

Melanie Mitchell, an AI researcher, mentioned an arXiv pre-print “Does GPT-4 Pass the Turing Test?”. Extracting this mention is already useful, and services like Altmetric collect these mentions as a simple proxy for attention over research. But we can go further with NLP, and try to parse this post into a nanopublication template which contains richer semantics. In this case, there is a template for “Commenting on or evaluating a paper”. For demonstration purposes, I played the role of an NLP model and manually parsed the tweet into that template:

This new form with semantics is much more useful than the raw natural language tweet!

Semantics are machine readable, enabling downstream apps to do all kinds of useful operations. For just one example, consider the landing page for papers, like you can find on arXiv or bioRxiv as shown below:

Semantic Nanopublications provide a richer context layer compared with existing social media mention counters

These landing pages often track social media mentions of research (using services like Altmetric as mentioned above), but since those posts lack semantics, all that can be done is a simple count display tracking how many times the research was mentioned. As a result, much information is lost, limiting the value of such signals. Now imagine we had semantics. We could display how each mention referred to the research in question, providing a much richer context for those who want to explore the discussion about a particular paper (please excuse the crude mockup, this is just for demonstration purposes!). This is just one example of many apps enabled by semantic publishing: more on that in future posts!

Another important contrast with services like Altmetric is that of consent. These services collect researchers' data without their consent or knowledge. Some may question whether consent matters in this case, given that tracking is only being conducted around scientific discourse, and social media data is public anyway. However, our mission is different, which is why we think consent matters: where existing services are focused on passive tracking of researchers' behavior (that gets creepy fast), Sensemaking Networks are aimed at developing active, participatory and ultimately more useful discourse around research (more here on the transition from "users to [sense]makers"). In this new model, data sharing and collection are intentional practices employed by communities and for communities to support their collective sensemaking efforts. Accordingly, we plan to experiment with different methods enabling researchers to have more control over the data they contribute. An early interesting example we've been inspired by is the Science Feed on Bluesky. As shown below, researchers can register themselves to the feed which then aggregates all of their posts marked with a 🧪 emoji:

Science Feed on Bluesky. Demonstrating a more participatory kind of feed

However, even if we address consent issues, converting social media posts to nanopublications using NLP is not a cure-all. While it can be extremely useful for bootstrapping the network, it is still only a partial solution: for one, AI tools are still far from perfect, despite truly impressive recent progress. In addition, many platforms have closed off their APIs completely or are making it increasingly costly to extract data. As observed by Barend Mons, one of the leaders of FAIR data and nanopublications, “Text mining? ...Why bury it first and then mine it again?”

That’s why we’re also planning to build a nanopublishing social app for researchers to use!

2. Building a decentralized semantic nanopublishing social app

There are a number of reasons why using an app for semantic publishing often makes more sense than writing unstructured natural language posts on social media, and hoping some AI system in the future will be able to make sense of it:

  1. Semantic publishing produces machine-actionable data. Unlike text mining, semantic publishing can instantly and directly power downstream applications. For example, sharing a review on Goodreads is a form of semantic publishing: the structured data is the review, which the app (Goodreads) can then process in useful ways (computing aggregate ratings for books, showing your friends updates, etc). However, in this example, the published data is also enclosed by Goodreads, meaning that only Goodreads can directly benefit from your data.

  2. Preserving data accessibility. Sensemaking Networks are designed to be resistant to platform enclosure, by building on top of open and decentralized social networking protocols like ATProto and ActivityPub. In an open ecosystem, the same data could play a role in multiple apps in parallel. For example, your social graph from Mastodon could also be used to filter reviews on an open & decentralized "Science Goodreads," instead of the situation today where each new platform forces you to create and maintain a new social graph. This transformative vision of radical interoperability and decoupling of apps from data is certainly not new, and indeed was the dream of early "Web 1.0" Semantic Web pioneers like Tim Berners-Lee. See below for some more thoughts on why the time might be right to revisit these exciting ideas.

  3. Preserving human authority. Finally, the very idea of AI automating the recovery of semantics from natural language without human supervision stands in tension to the principle of source authoritativeness in semantic publishing:

    "Semantic representations can only be considered authentic if they originate from an agent that is authoritative in the given situation. In the case of the publication of a scientific result, the only authoritative source are the researchers (who are called authors in this context). Semantic representations of scientific results are only authentic if they are provided by the researchers themselves, and this relation can be made explicit with a precise provenance representation." -- Kuhn & Dumontier (2017). Genuine semantic publishing

    This argument is especially relevant in an age of rampant AI-generated content. Tracking the provenance of information and being able to discern what was generated by humans vs AI agents will be increasingly important. In many cases, AI might assist researchers in generating content, but humans must remain the ultimate arbiters: "any attribution of authorship carries with it accountability for the work, and AI tools cannot take such responsibility."

So what would a semantic nanopublishing app look and feel like?

One simple option is: just like Mastodon, Bluesky, or equivalent apps, but with the key difference that instead of publishing just posts, users can nanopublish, meaning that they can select a template best expressing the semantics of their message.

From posting to semantic nanopublishing

Besides appearing in other people's feeds, radical interoperability means these nanopublications can also easily flow into other apps and services. Continuing the example of the bioRxiv landing page from Section 1 above, a researcher's nanopublished assessment of some paper could directly appear in context on the landing page of that paper, in addition to appearing in social media feeds. Similarly, inline annotations (like those made using hypothes.is) could "live" simultaneously on social media and as a layer over web content:

Semantic publishing needs to be as easy and natural as posting on existing apps, and great UX + AI assistants will be key to making that possible. For example, users can start typing their post in natural language and an interactive AI assistant can suggest suitable templates. More broadly, one of the reasons that early versions of the Semantic Web didn't pan out as expected was that they introduced additional friction and technical complexity that people were not prepared to handle. We're excited by the promise of revisiting semantic publishing with the benefit of years of progress in AI, interface design and distributed computing technologies!

Cooperative ecosystems, not platforms

Building great infrastructure is a huge undertaking, and a fuller realization of Sensemaking Networks is ultimately beyond the scope of a one year residency. Also, our goal is not to build yet another social media platform! Rather, evolving, interoperable and collaborative ecosystems are both the means and ends of the project. Therefore, we're building partnerships with mission-aligned organizations to create this new ecosystem, together. Shout-outs to some awesome groups we've already started working with (hopefully lots more to come soon!):

And of course, we're very excited to continue building our collaboration within Astera and its network.

As usual, thoughts and comments are very welcome - feel free to reach out at ronen@astera.org

That's all for now! Hopefully next time we'll have some actual screens or prototypes for you 🤞

Happy 2024, make merry and make sense! 🔄🍾🎉

]]> cosmik@newsletter.paragraph.com (Ronen Tamari) <![CDATA[Sensemaking Networks: Project Introduction]]> https://archive.cosmik.network/sense-nets-intro RoLMm8IJmfPpD9e00Sxp Wed, 13 Dec 2023 00:00:00 GMT I’m excited to start this blog as I begin a new chapter as an Entrepreneur-In-Residence (EIR) at the Astera Institute! Astera is launching the visionary EIR program to catalyze the upgrading of our aging infrastructure for science publishing and communication.

In this post I'll tell you about my planned residency project, which I'll be developing on along with partners at Astera, Common Sensemakers and beyond (we're still in the process of team formation).

The project is called Sensemaking Networks, an upgrade of social media designed for scientists and other knowledge workers, combining the strengths of AI-augmented social media networks with decentralized semantic nanopublishing, to radically enhance information flow as well as participants’ agency and data sovereignty.

This first post will focus more on the "why" of the project (background and motivation), and in future posts we'll focus more on the "what" and "how" of implementation.

Sensemaking as a missing layer from the scientific record

As we and others have previously made the case for, science social media is far from just an idle diversion for scientists while they aren’t writing papers. Rather, it merits treatment as an integral part of the scientific process: science social media is part of the sensemaking layer of science that helps us to collaboratively communicate, contextualize, curate and assess research. The sensemaking layer improves the inclusivity of science, bringing the broader public into direct contact and participation in discourse around science. However, this blurring of the lines between scientists and "citizen scientists" is unfortunately part of the reason this layer gets ignored. Another reason is the "messiness" of the sensemaking layer, which can include any kind of knowledge shared by people as they engage with scientific research, such as annotations, reviews and social media commentary. More generally, we tend to ignore what's hard to measure, and sensemaking is a complex process where knowledge hasn’t yet crystallized into a polished paper:

And yet, we ignore the sensemaking layer at our peril: Science has not taken ownership of the sensemaking layer, and as a result, it is siloed and fragmented across many platforms, with profiteering publishers vying for control of researchers’ data as part of their transformation into data tracking and analytics conglomerates.

The sensemaking layer (blue): non-traditional publications related to science research, like posts on science social media. These are currently "dark matter" in the scientific knowledge universe (left), which we aim to bring into the open using Sensemaking Networks (right)

These corporations’ operations are often at odds with basic scientific values such as transparency and free access to knowledge. More importantly perhaps, researchers are overwhelmed with information, yet lack access to the people, tools and data that will help them collectively make sense of research and advance science.

So what are the problems with the existing infrastructure, and how does our project fit into the landscape of existing solutions?

Limitations of existing science communication infrastructure

Sensemaking Networks are designed to address three interrelated limitations of traditional (and non-traditional) science publishing and communication:

  1. Lack of support for diverse publishing formats and scales.

  2. Feedback loops are too long and opaque, and publications’ reach is poor.

  3. Non-traditional publications are siloed and fragmented across many apps and formats.

Let's look at each of these in detail:

  1. Lack of support for diverse publishing formats and scales. In traditional publishing, the basic unit of knowledge is a paper, a format that has changed surprisingly little over the past hundreds of years, given that since papers, we’ve also invented computers and the internet. Thankfully, in recent years people have been expanding the concept of publishing to include a far wider range of activities, both in terms of formats as well as scale. The Nanopublications project is a great example of this nascent trend: as the name implies, nanopublications are aimed at representing very small units of knowledge - typically corresponding to a sentence or two in natural language. What makes nanopublications especially interesting is that they employ Semantic Web principles to represent publications as small knowledge graphs, which are great for representing information in an unambiguous, machine-actionable manner. Nanopublications are also diverse: they can express virtually any kind of knowledge, from a relation between a gene and a disease, to an opinion about a scientific paper or a blog post. More later on why nanopublications are so cool and potentially transformative.

Example nanopublication! (source) A nanopublication is like a social media post, but with semantics (structured fields defining relations to other informational entities).

  1. Feedback loops are too long and opaque, and publications’ reach is poor. Traditional publishing is a slow back-and-forth between authors and reviewers, taking months or even years from submission to publication. The review itself is usually hidden from the public, creating a process that is opaque (other researchers in the field could have learned from the review) and non-inclusive (only 3 randomly selected reviewers can participate, why not others?). Finally, traditional methods do little to improve the reach of publications; most research is lost in the deluge of new papers, perhaps never to be found by its intended audience. These shortcomings of traditional publishing help explain why science social media (shorthand for the use of social media for science communication) has become so popular: scientists can disseminate their research on social media (e.g., Twitter, Bluesky, etc) and receive immediate feedback. Social media’s public nature invites broad participation, and algorithmic feeds coupled with social graphs greatly improve reach: retweets by “science influencers” can lead to millions of views. A key insight we'll be returning to throughout is that, in many ways, researchers are already nanopublishing on social media, but just don’t know it yet. What if Twitter was real academic work?

  2. Non-traditional publications are siloed and fragmented across many apps and formats. While the Open Science movement has significantly improved access to traditional publications (papers), non-traditional publication is a wild west of non-interoperating apps and formats. As noted above, science social media functions de facto as an important nanopublishing venue where all manner of valuable knowledge is shared. In many cases, commercial platforms like Twitter enclose this data as part of their business model, directly hindering open science. Even on more open social platforms such as Mastodon or Bluesky, the lack of standard formats and support for semantic publishing makes it difficult to effectively aggregate, filter and index data at scale and across platforms.

Enter Sensemaking Networks

We envision Sensemaking Networks as federated/decentralized Twitter-style social networks where posts are nanopublications, and data is open and FAIR (Findable, Accessible, Interoperable and Reusable). As shown in the Venn diagram below, our approach aims to draw on the best of social media networks in terms of feedback and reach, while addressing their limitations of siloed data and restricted publishing expressivity.

Sensemaking Networks in the context of current challenges in science publishing

Nanopublications are great for expressing diverse kinds of knowledge, but can be clunky to create and lack the “aliveness” of social networks. By embedding nanopublications in social networks we can enhance their discoverability and encourage their broader adoption by researchers who want to not only publish, but also engage with other researchers and the broader public. AI will play multiple key roles in Sensemaking Networks, both to make semantic publishing as easy as posting on regular social media, and also for improving reach and feedback, for example by connecting people with relevant content and other researchers.

In the next post we'll dive into our project plan and more concrete implementation details, stay tuned! Please subscribe to this blog if you want to stay updated about project developments, and potential contribution or job opportunities. Also feel free to reach out at ronen@astera.org if you want to chat about this more!

That's it for now! 🚀

]]> cosmik@newsletter.paragraph.com (Ronen Tamari)