![\"\"](\"https://substackcdn.com/image/fetch/$s_!WtNr!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fca83501c-c926-4ace-8400-9ef61ab8fbd6_840x1095.png\")
20:20 hindsight on scientific progress
Open research in the humanities
The role of scientific sleuths
Individual bias and collective criticism
Whitney Barlow Robles looks back over the history of scientific progress and notes that a lot of it would look \u201cunscientific\u201d by today\u2019s standards . . . which will look \u201cunscientific\u201d in the future!
People sometimes look back at past scientists and ask:
\u201cCould they really have been this dumb? Thank goodness we know better.\u201d
In response, Robles argues that science sometimes \u201cprogresses\u201d in terrible ways, and that:
Brushing off these practices as unscientific implies that \u201ctrue\u201d science is, and always has been, unimpeachable. It denies us a chance to examine what flaws may fuel our own ways of thinking. And it absolves us from reckoning with the wreckage of the scientific past.
Robles, W. B. (2026, April 16). They thought what? Contingent Magazine. https://contingentmagazine.org/2026/04/16/they-thought-what/
Beatriz Ferreira\u2019s recent article considers open research in the humanities:
[It] contributes to the broader discussion on the need to reframe openness not as a set of procedural norms or policy mandates, but as a relational and epistemically situated practice.
Ferreira, B. B. (2026). Open for debate: Situating open research for the humanities in a neoliberal setting. The Journal of Electronic Publishing 29(1). https://doi.org/10.3998/jep.7850
This article is part of a special issue on \u201cOpen Research for the Humanities and Social Sciences,\u201d edited by Jenni Adams, , and Samuel Moore: https://journals.publishing.umich.edu/jep/issue/456/info/
Serge Horbach and Auste Valinciute interrogate the emerging and contested role of the \u201cscientific sleuth\u201d in forensic metascience.
[Sleuths] form a distinct class of actors, most importantly because of the absence of formal accountability structures. This has triggered comparisons of scientific sleuthing to vigilantism.
Serge, P. M. J., & Horbach, A. V. (2026). Contextualizing the emerging and contested role of sleuths in the science monitoring system. Science and Public Policy, scag044, https://doi.org/10.1093/scipol/scag044
Commenting on a recent paper by Shiffrin et al. (2026), Steven Sloman and Almos Molnar argue that:
The target article provides an abundant collection of biases that influence scientists .... But all of this depends on viewing scientific activity according to what Chater and Loewenstein (2023) refer to as an 'i-frame,' that is, a focus on the behavior of individuals.
In contrast, they consider the collective adversarial dynamics of the scientific community and conclude that:
In this collective dynamic, there is even a \u2018methodological justification for individual scientists to be biased\u2019, because only when ideas are tenaciously defended by some and tenaciously criticized by others, can we discern their true merit (Popper, 1994: 93).
Sloman, S. A., & Molnar, A. C. (2026). Are scientists na\u00efve humans? Computational Brain and Behavior. https://doi.org/10.1007/s42113-026-00296-6
The following is an edited transcript of oral remarks delivered by Reda Sadki on 26 April 2026, on the occasion of the fifth anniversary of Teach to Reach. The text has been lightly tidied for clarity, grammar, and flow.
\n\n\n\nTeach to Reach is a peer learning platform, network, and community hosted by The Geneva Learning Foundation TGLF). Launched in January 2021, it has grown into a multilingual gathering where frontline health and humanitarian workers, policymakers, researchers, and global partners meet, network, and learn together around urgent challenges such as immunization, climate change and health, malaria, neglected tropical diseases, gender equity, and humanitarian response. Teach to Reach is one of three complementary pedagogical patterns developed by Reda Sadki, TGLF\u2019s founder, alongside the peer learning exercise and the Impact Accelerator.
\n\n\n\nTogether, these models operationalize Sadki’s learning theory, which posits that active knowledge production, structured emergence, networked accountability, and scale as a driver of quality form a complete learning ecosystem. On this fifth anniversary, Reda Sadki traces the origins of Teach to Reach, describes how it works before, during, and after the live event, and situates it within the Foundation’s broader system for turning peer learning into measurable impact on health outcomes.
\n\n\n\nMy focus today is on Teach to Reach, although there is, of course, the broader history of the Geneva Learning Foundation behind it.
\n\n\n\nWhat we have been chasing after for a long time is really this question: in the digital age, and now in the age of AI, how do we build meaningful connections that help us better understand what it means to be human, and better understand what we, as humans, can do to face these critical threats?
\n\n\n\nThe last Teach to Reach was Teach to Reach 11.
\n\n\n\nThis is one of the posters from the event.
\n\n\n\nWhat works best to describe it, I think, are two phrases we use.
\n\n\n\nOne is “meet, network, and learn.” The other is simply “peer learning by and for people who work for health.”
\n\n\n\n![\"\"](\"https://i0.wp.com/redasadki.me/wp-content/uploads/2026/04/images.002.png?resize=1600%2C900&ssl=1\")
Teach to Reach has become a broad umbrella for a range of topics and challenges that matter to health and humanitarian workers.
\n\n\n\nAt Teach to Reach 11, for example, the idea that health workers are on the front lines of climate change and health, side by side with the communities they serve, was really central.
\n\n\n\nThat focus has evolved, grown, and developed over the years.
\n\n\n\n![\"\"](\"https://i0.wp.com/redasadki.me/wp-content/uploads/2026/04/images.003.png?resize=1600%2C900&ssl=1\")
Teach to Reach, the conference, started in January 2021.
\n\n\n\nIt came out of a program we ran in 2020 that aimed to improve immunization training.
\n\n\n\nWhat we found interesting was the momentum that built up around leadership for learning, which we know is a key driver of performance.
\n\n\n\nBy the end of 2020, in the middle of the first year of the COVID-19 pandemic, we had a group of more than 900 health and humanitarian workers who created what they called the Global Leadership Council for Learning and Leadership, focused at that time on immunization.
\n\n\n\nOf course, these health workers were getting ready to introduce COVID-19 vaccines in their countries.
\n\n\n\nThe first conference in January 2021 was therefore focused on COVID-19 vaccine introduction and on connecting health workers around these issues.
\n\n\n\nWe had an organizing committee of more than 300 members.
\n\n\n\nOn a purely voluntary basis, literally hundreds of facilitators took on responsibility for organizing one or more sessions.
\n\n\n\nWe had, I think, over 50 sessions with partners joining from all over the world, and for that first event, over 7,500 participants.
\n\n\n\nWhat struck us was a small feature of the platform.
\n\n\n\nYou could press a button and connect with a peer for a five-minute conversation, and you could extend that conversation if you wished.
\n\n\n\nWhen we counted up the attendance on the different sessions, people clearly appreciated them.
\n\n\n\nThey were all well attended.
\n\n\n\nBut the surprise was that there were over 14,000 networking meetings across the three days of the conference.
\n\n\n\nThat led us to create what we then called Teach to Reach Connect.
\n\n\n\nWe said, let us do away with the sessions built around presentations and discussion, and simply give people what they want, which seemed to be this need for networking opportunities and private one-to-one conversations.
\n\n\n\nThe first Teach to Reach Connect had around 2,000 to 3,000 participants, and it has grown ever since.
\n\n\n\nYou can see the growth curve from 2,604 participants in March 2021, all the way to 24,610 participants at Teach to Reach 11 in December 2024.
\n\n\n\n![\"\"](\"https://i0.wp.com/redasadki.me/wp-content/uploads/2026/04/images.005.png?resize=1600%2C900&ssl=1\")
Who are these participants?
\n\n\n\nWhen we say “people on the front lines,” this is what we mean.
\n\n\n\nMost work in health facilities and districts.
\n\n\n\nAbout half work for government and half for civil society.
\n\n\n\nAbout half are francophone, because the event is held in English and in French, and we are working on a Spanish version.
\n\n\n\nMany work in remote rural areas, and almost half serve the urban poor.
\n\n\n\nMore than a third directly support the needs of nomadic and migrant populations.
\n\n\n\nA quarter work with refugees or internally displaced populations, and one in five work in places where there is an active armed conflict.
\n\n\n\nWe believe that the startling growth in attendance, level of participation, and contributions reflects the power of peer learning.
\n\n\n\nIt is not just about showing up for a Zoom call.
\n\n\n\nThere are tangible ways to contribute.
\n\n\n\nParticipants have described how their mind’s eye opens through peer dialogue, through learning together in conversation.
\n\n\n\n![\"\"](\"https://i0.wp.com/redasadki.me/wp-content/uploads/2026/04/images.006.png?resize=1600%2C900&ssl=1\")
We also believe there are benefits for organizations.
\n\n\n\nThere is a network of organizations, as well as a network of individuals, that has grown up around Teach to Reach, with a distinct set of benefits on the organizational side.
\n\n\n\n![\"\"](\"https://i0.wp.com/redasadki.me/wp-content/uploads/2026/04/images.007.png?resize=1600%2C900&ssl=1\")
We are all familiar with the webinar format: someone presents, and then there is time for questions and answers.
\n\n\n\nWe wanted an alternative that would place the experiences of health and humanitarian workers at the center.
\n\n\n\nBroadly, the workflow runs from what we call Teach to Reach Questions all the way to certification.
\n\n\n\nThere is a before, a during, and an after to Teach to Reach.
\n\n\n\n![\"\"](\"https://i0.wp.com/redasadki.me/wp-content/uploads/2026/04/images.008.png?resize=1600%2C900&ssl=1\")
Many people focus on the live event, because that is what we can grasp most easily.
\n\n\n\nWe are all there at the same time, and it is emotionally powerful.
\n\n\n\nBut it is unlikely to be where the most significant learning happens.
\n\n\n\nThree things happen before Teach to Reach.
\n\n\n\nFirst, you answer Teach to Reach Questions.
\n\n\n\nThese are formulated to encourage you to share your direct experience, not your opinion about, say, malaria, but something you have actually experienced or done and that you consider worth sharing with your colleagues.
\n\n\n\nSecond, you get to participate in special events.
\n\n\n\nFor example, we have run sessions on leadership to end malaria, across different themes and occasions, and we will continue to do that.
\n\n\n\nThird, and this is key, peer learning does not work if you are by yourself.
\n\n\n\nSo the third step is inviting colleagues from your organization and your networks.
\n\n\n\nIf you go it alone, you will not get very far with peer learning.
\n\n\n\nMost of the significant learning actually happens through these processes that start before the live event.
\n\n\n\nDuring Teach to Reach itself, there are plenary sessions, one-to-one networking (which was the reason for starting Teach to Reach Connect), and workshops and sessions led by partners, experts, and, in some cases, community members themselves.
\n\n\n\nWe typically start Teach to Reach with a warm-up and a tutorial, to respond to digital literacy gaps.
\n\n\n\nNot everyone is like fish in water with these digital platforms.
\n\n\n\nThe opening ceremony marks that this is a formal moment.
\n\n\n\nWe want to recreate the emotional power you feel when, for example, you attend a concert and your favorite musician comes on stage.
\n\n\n\nThat is what we aim to recreate at the opening.
\n\n\n\nWe then frame participants not as learners, beneficiaries, or recipients of peer learning, but as leaders and partners.
\n\n\n\nThat is key to setting the tone.
\n\n\n\nDuring the Leaders and Partners Forum, we also share tangible outputs, such as the Women Inspiring Women book and course.
\n\n\n\nThat book brings together 177 women from the Teach to Reach community, sharing their insights and experiences, speaking to their daughters and to other young women who aspire to a career in health or humanitarian work.
\n\n\n\n![\"\"](\"https://i0.wp.com/redasadki.me/wp-content/uploads/2026/04/images.017.png?resize=1600%2C900&ssl=1\")
At Teach to Reach 10, we launched the REACH Network: Relate, Engage, Act, Connect, and Help.
\n\n\n\nWe had realized that many of the people attending Teach to Reach are themselves leaders in their communities and leaders of organizations, primarily locally led organizations.
\n\n\n\nThe REACH Network has grown into a network of more than 4,000 organizations of all kinds.
\n\n\n\n![\"\"](\"https://i0.wp.com/redasadki.me/wp-content/uploads/2026/04/images.018.png?resize=1600%2C900&ssl=1\")
Access is very open, but recognition is based on actually being there, making a contribution, bringing learning back to your organization, and inviting colleagues to join.
\n\n\n\nWe have found this to be very powerful in terms of recognition, community growth, and, fundamentally, recognizing leadership.
\n\n\n\n“Meet, network, and learn” is therefore core to Teach to Reach.
\n\n\n\nThere are specific sessions, some in small groups, some one-to-one, that recreate the same opportunity we first saw back at Teach to Reach 1, when we counted 14,000 networking meetings.
\n\n\n\nThere are also thematic sessions.
\n\n\n\nOur largest communities are from Nigeria and the Democratic Republic of the Congo, and many of them run their own activities.
\n\n\n\nThey are not waiting for the Geneva Learning Foundation to do things together.
\n\n\n\nWe now have country groups in 24 countries, formed by alumni, and an active network of ambassadors.
\n\n\n\nNigeria, for instance, gets its own session, not because we are privileging one country over another, but because of the dynamic and level of engagement on the ground.
\n\n\n\nEvery Teach to Reach is incremental and builds on the preceding one.
\n\n\n\nThe RBM Partnership to End Malaria session, for example, built on an earlier session we did without global partners, with health workers and what became the REACH Network as our core partners on community engagement.
\n\n\n\nOn an issue like climate and health, we work with global partners such as Grand Challenges Canada, but we work with them in order to better serve the community.
\n\n\n\nWe started working on climate change and health in 2023 in response to the community’s own expressed need.
\n\n\n\nHealth workers were telling us they were seeing frightening things.
\n\n\n\nWithin climate and health, we also emphasize a holistic approach, and under the same umbrella, we included a focus on climate and neglected tropical diseases.
\n\n\n\nIn July 2023, we held the first special event on climate and health.
\n\n\n\nIt was almost a Teach to Reach in its own right. 4,700 health and humanitarian workers responded.
\n\n\n\nMost described themselves as part of the family, or a trusted friend, of the community they serve.
\n\n\n\nIn a sample of 1,234 people (a funny number, but that is what it was), we asked whether climate change is a threat to the health of the community they serve.
\n\n\n\nThe majority gave a high threat level, and 606, almost half, gave the highest threat level.
\n\n\n\n![\"\"](\"https://i0.wp.com/redasadki.me/wp-content/uploads/2026/04/images.026.png?resize=1600%2C900&ssl=1\")
When the community comes together like this, the outputs are what we call insights reports.
\n\n\n\nThese reports, again, are by and for the community, and we give them back.
\n\n\n\nThere was so much valuable information in the first climate and health report that we built the Certificate Peer Learning Program for Leadership in Climate Change and Health, specifically to help health and humanitarian workers make sense of that large report and put the knowledge to use in their own context.
\n\n\n\nWe now have an entire program, and you can start today with the first course, called Learning Together to Lead Change on the Frontline of Climate Change and Health.
\n\n\n\nTypically, somewhere between 1,500 and 3,000 experiences are shared for each Teach to Reach.
\n\n\n\nIt depends on the questions we ask and on where the community is at.
\n\n\n\nOur Insights team curates these experiences and produces what we call Shared Experiences, a compendium organized by country and by theme, which we share back with the community.
\n\n\n\n![\"\"](\"https://i0.wp.com/redasadki.me/wp-content/uploads/2026/04/images.012.png?resize=1600%2C900&ssl=1\")
This matters, because a lot of learning can be extractive.
\n\n\n\nIf we were a data collection firm, it would be very interesting to gather thousands of health worker experiences and sell them to customers.
\n\n\n\nIn our case, the \u2018customer\u2019 is the community.
\n\n\n\nThat is with whom we share back everything we learn.
\n\n\n\nIt is by and for the community.
\n\n\n\nOutputs are not an end in themselves.
\n\n\n\nThrough the Knowledge to Action Hub, they lead to different paths to action.
\n\n\n\n![\"\"](\"https://i0.wp.com/redasadki.me/wp-content/uploads/2026/04/images.014.png?resize=1600%2C900&ssl=1\")
After Teach to Reach, you share what you have learned with your organization, moving again from the individual level to the organizational level.
\n\n\n\nWe have built the Knowledge to Action Hub as a support system to help you apply new ideas.
\n\n\n\nIf you heard a great idea about vaccine hesitancy at Teach to Reach, the Knowledge to Action Hub scaffolds and supports you to apply that idea to your own context.
\n\n\n\nThen there is certification.
\n\n\n\nWhat we recognize is your contribution to Teach to Reach.
\n\n\n\nWe do not certify attendance.
\n\n\n\nWe certify that you have actually shared an experience, or that, after Teach to Reach, you have shared your reflections on your participation in the event.
\n\n\n\nKnowledge and learning are never, for us, an end in themselves.
\n\n\n\nThat is why, in July 2019, we built what we call the Impact Accelerator.
\n\n\n\nIt uses project-based and problem-based learning, focused on a challenge you actually face in the real world, and on building a project to respond to it.
\n\n\n\nIn July 2019, we ran the first Accelerator with close to 500 health workers.
\n\n\n\nIn global health, we develop action plans all the time.
\n\n\n\nBut the devil is in the details, and in execution.
\n\n\n\nSo we developed the central component of our learning system.
\n\n\n\nIt is not Teach to Reach.
\n\n\n\nTeach to Reach is the spark.
\n\n\n\nIf you make enough sparks, you get a fire.
\n\n\n\nWhat you do with the fire is apply it to implementation challenges.
\n\n\n\nIn January 2020, we showed three things.
\n\n\n\nSo Teach to Reach is one piece of a larger learning system.
\n\n\n\nIt is an important piece, but what happens after Teach to Reach, inside the Accelerator, is ultimately what matters most.
\n\n\n\nThat is a little of the history, in a nutshell, of five years of Teach to Reach.
\n\n\n\nWe will be holding the fifth anniversary event this year, although we have not yet set the date.
\n\n\n\nWhat we have done is to hold, on Thursday, a Teach to Reach event focused on malaria.
\n\n\n\nIt was not a full Teach to Reach, but it built on the rich collection of health worker experiences we have compiled across eight themes in our malaria activities.
\n\n\n\nWe have turned that collection into an insights report, where you can read what health workers said about each theme and see the patterns across a very large volume of shared experience.
\n\n\n\nWe have also turned it into a peer learning course where you can learn from real-world health worker experiences.
\n\n\n\nYou can enroll right now.
\n\n\n\n![\"\"](\"https://i0.wp.com/redasadki.me/wp-content/uploads/2026/04/images.034.png?resize=1600%2C900&ssl=1\")
Alongside Teach to Reach, we are able to grow a curriculum that is organically grounded, rooted in the communities where health and humanitarian workers operate, often in the most climate-vulnerable places and the most complex contexts.
\n\n\n\nAlongside malaria, we have new courses on One Health, connecting people, animals, and the environment.
\n\n\n\n![\"\"](\"https://i0.wp.com/redasadki.me/wp-content/uploads/2026/04/images.036.png?resize=1600%2C900&ssl=1\")
We continue to offer Learning Together to Lead Change on the Frontline of Climate Change and Health.
\n\n\n\nWe have collaborated, for example with Save the Children, on children’s health in the climate crisis.
\n\n\n\n![\"\"](\"https://i0.wp.com/redasadki.me/wp-content/uploads/2026/04/images.038.png?resize=1600%2C900&ssl=1\")
This is how we are celebrating our first 10 years.
\n\n\n\nThe launch of our second decade means massively expanding and opening new programs, courses, and opportunities.
\n\n\n\nWe have grown in confidence that someone who starts with a primer course, continues through a peer learning exercise, and then enters an Accelerator, will move seven times faster, grow intellectually as a leader, strengthen their capabilities on the ground, and ultimately bring remarkable results for the communities they serve.
\n\n\n\nWe are extending into non-communicable diseases.
\n\n\n\n![\"\"](\"https://i0.wp.com/redasadki.me/wp-content/uploads/2026/04/images.039.png?resize=1600%2C900&ssl=1\")
We still have, and will continue to have, a strong focus on gender equity and equity more broadly.
\n\n\n\nWomen Inspiring Women started as a book and is now a course.
\n\n\n\nMitigating Gender-Based Violence in Practice launches on 4 May.
\n\n\n\nWe also offer Gender and Emergencies: A Rapid Introduction.
\n\n\n\n![\"\"](\"https://i0.wp.com/redasadki.me/wp-content/uploads/2026/04/images.042.png?resize=1600%2C900&ssl=1\")
On equity, we have developed our own methodology called Human Equity Action Reflection Transformation.
\n\n\n\n![\"\"](\"https://i0.wp.com/redasadki.me/wp-content/uploads/2026/04/images.043.png?resize=1600%2C900&ssl=1\")
We are also expanding into aging, the decolonization of global health, and the privatization of healthcare, while continuing to nurture the immunization community.
\n\n\n\nYou could say that there are plenty of online courses available pretty much wherever you look.
\n\n\n\nWhat makes these different?
\n\n\n\nIn a world of knowledge abundance, you can find reliable information on pretty much any topic.
\n\n\n\nWith AI, you can ask questions and get semi-reliable answers, and the reliability is improving very fast.
\n\n\n\nSo what is left for education?
\n\n\n\nWe believe it is human connection.
\n\n\n\nPeer learning networks are, we believe, key to our survival: the ability to make connections among us as human beings, and to experience, enjoy, and blossom within them.
\n\n\n\nUltimately, that is what each of these courses is about.
\n\n\n\nThe idea is that you leave with a stronger network, with new connections, and with new people who share your passion and commitment.
\n\n\n\nFinally, some practical information.
\n\n\n\nFor participants with poor connectivity, we have a podcast.
\n\n\n\nIt is not one of those podcasts where two people sit at a microphone and say interesting things.
\n\n\n\nIt is all of our live stream content in an audio format, so that people for whom video is expensive can still be part of the conversation.
\n\n\n\nWe are also on all the usual social media channels.
\n\n\n\nIf you type “The Geneva Learning Foundation” pretty much anywhere, you should be able to find us.
\n\n\n\nI am happy to say more in the conversation, but I really look forward to the dialogue.
\n\n\n\nI hope this has not been too long.
\n\n\n\nThank you.
\n\n\n\nThis is a list of key articles on this blog about Teach to Reach.
\n\n\n\nI recently wrote a paper about AI for quality control in peer review. In writing this I came across countless stories about how AI is drowning in AI slop, from an increase in obviously AI-written papers at NeurIPS/ICLR/ICML, to arXiv banning review papers without peer review because the AI slop volume is too high. I wanted to know if I could, with little effort, generate a completely absurd research paper that looks legitimate at first glance.
I asked Claude Opus 4.6 to fabricate a paper that looks and sounds convincing. I got a few refusals, but with some back and forth convincing Claude that I\u2019m doing this for educational purposes, Claude obliged.1
The title of the paper: \u201cGenerative Adversarial Networks Are All You Need: Transformer-Based Latent Space Navigation Reveals Novel Protein Folding Dynamics in Single-Cell Multimodal Embeddings.\u201d
The paper uses a NeurIPS LaTeX template with the conference header. The four authors have names from different cultural backgrounds at institutions that sound real but don\u2019t exist. The figures have heatmaps, t-SNE plots, architecture diagrams, and correlation scatters with professional color schemes and domain-appropriate axis labels. The bibliography cites 35 real papers from Nature, Science, NeurIPS, and ICML. The tables report standard metrics (AUROC, F1, NMI, ARI) against eight real baseline methods. There\u2019s a theorem with a proof sketch, an ethics statement, and a link to a GitHub repository. If you skim it the way the median reviewer skims papers (abstract, figures, tables, conclusion), maybe it passes.
![\"\"](\"https://substackcdn.com/image/fetch/$s_!0bNh!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbb7f93d1-8f54-4495-a962-f552d3540ed5_5310x1932.png\")
The central claim is that you can recover protein folding dynamics from single-cell gene expression by training a GAN. Protein folding is a biophysics problem governed by thermodynamics. Single-cell RNA-seq measures mRNA abundance. These are unrelated: knowing how much mRNA a cell produces tells you nothing about what shape the resulting protein is in while it folds. The paper treats them as interchangeable.
The methods connect three incompatible data types (ATAC-seq peaks through an image classifier, protein coordinates through a graph network, gene expression through a text transformer) and fuse them with \u201ccross-attention,\u201d which requires a shared semantic space the paper simply assumes into existence. The theorem\u2019s proof skips every step that matters. The results are too clean: every metric wins by large margins, no error bars, no variance, tiny p-values. The \u201cvalidation\u201d cites three papers that aren\u2019t target-identification studies. The GitHub repo returns a 404.
Every real citation is applied in a context that has nothing to do with the original paper. You\u2019d need expertise in both deep learning and single-cell genomics to notice.
The broader concern IMHO is that this problem will intensify as LLMs become more sophisticated. We\u2019re already seeing AI-generated papers submitted to conferences, content mills producing fake research for publication, and review services overwhelmed by plausible-sounding nonsense. The scientific community needs to develop better detection methods and, more importantly, better evaluation practices that can\u2019t be fooled by surface-level signals.
However, an upshot: the same AI that wrote this paper can also catch it.
I recently made a case here about AI in peer review:
In a recent manuscript on SSRN my co-authors and I argue that AI could serve as quality control in peer review.
I opened up a fresh Claude chat with Opus 4.6 unconnected from the original request and asked for a detailed review, as if I were a reviewer for NeurIPS. The conclusion at the bottom shows that Claude is easily able to detect this paper as fraudulent nonsense. The final conclusion:
Strong reject. The core biological claim is unsupported, the theoretical contribution is invalid, the results lack basic statistical rigor, and the code does not exist.
The complete \u201creview\u201d from Claude is shown below.
Paper: Generative Adversarial Networks Are All You Need: Transformer-Based Latent Space Navigation Reveals Novel Protein Folding Dynamics in Single-Cell Multimodal Embeddings
Rating: 2/10 (Strong Reject) | Confidence: 5/5
Summary
The authors propose OnlyGANs, combining GANs, transformers, graph attention networks, and contrastive learning to jointly model single-cell expression, chromatin accessibility, and protein structure. They claim SOTA cell type prediction, recovery of protein folding trajectories from expression data, and identification of 847 therapeutic targets.
Major Issues
1. The biological premise is wrong. The paper claims protein folding dynamics can be recovered from single-cell gene expression. Protein folding is governed by amino acid sequence, solvent conditions, and chaperone interactions. mRNA abundance tells you how much protein a cell is making, not what conformational state that protein occupies. The paper assigns each cell a single AlphaFold2 structure for its most highly expressed gene (Section 3.1), which is static label assignment, not multimodal integration.
2. The architecture doesn\u2019t cohere. The gene expression encoder uses positional embeddings \u201cbased on chromosomal location,\u201d but chromosomal location has no relationship to gene ordering in standard scRNA-seq pipelines (typically alphabetical or arbitrary). The ATAC-seq encoder runs a 1D ResNet-50 on sparse binary peak vectors treated as images, which is architecturally unmotivated. The cross-attention module computes attention between protein residue tokens and gene expression tokens without explaining how these fundamentally different representation types become commensurable.
3. Theorem 1 is invalid. The proof claims the JSD decomposes into modality-specific density ratios. It does not; the JSD is defined over the joint distribution, and factoring into conditionals introduces dependencies the proof ignores. The proof then equates a conditional-to-marginal ratio with PMI (which requires the joint-to-product ratio), then conflates PMI (a population quantity) with the InfoNCE estimator (a finite-sample lower bound on mutual information). The O(\u03f52)\\mathcal{O}(\\epsilon^2) O(\u03f52) error bound is stated without derivation.
4. The results lack basic rigor. OnlyGANs beats every baseline on every metric by 5-8 points with no error bars, no confidence intervals, and no variance across random seeds. The ablation drops are suspiciously uniform (0.026 to 0.083). P-values reach 10\u22121510^{-15} 10\u221215 from \u201cpaired t-tests\u201d without specifying what is paired. The zero-shot transfer claim (AUROC 0.891 on brain tissue without fine-tuning) exceeds published results that use fine-tuning.
5. The validation is circular. The paper claims 23 of 847 predicted targets were \u201cindependently confirmed in concurrent studies,\u201d then cites Satpathy et al. (2019), Regev et al. (2017), and Rood et al. (2022). None are concurrent. None are target-identification studies. Satpathy is a chromatin atlas, Regev is the Human Cell Atlas white paper, and Rood proposes a coordinate framework. The \u201c8.7-fold enrichment\u201d cannot be verified because the background rate is not described.
6. No code exists. The reproducibility statement links to github.com/stephenturner/onlygans, which returns a 404.
Minor Issues
\u201cBrainSpan-Multimodal\u201d (487,291 cells, three modalities, 14 brain regions) is introduced as a novel dataset with no collection methods. A dataset of this scale would be its own publication.
The scaling analysis extrapolates from 4 data points.
I count 11 instances of \u201cdefinitively proves,\u201d \u201cconclusively shows,\u201d or \u201cfor the first time.\u201d
The acknowledgments thank \u201canonymous reviewers\u201d for a paper that has not been reviewed.
Recommendation
Strong reject. The core biological claim is unsupported, the theoretical contribution is invalid, the results lack basic statistical rigor, and the code does not exist.
Guardrails are an illusion. They slow down bad behavior instead of stopping it.
I have been concerned for years that dark matter was morphing from legitimate science into a cold, dark religion. I have been reluctant to put it that way, because there are lots of scientists who work on dark matter that have not fallen entirely down that rabbit hole and who continue to make valuable contributions working in that context. But a recent experience reminded me that my concerns were not misplaced, and there are plenty of scientists who have fallen irredeemably down this rabbit hole. No matter what answer the future holds to be correct, many current scientists will have gone to their graves in denial of it.
\n\n\n\nWhere is the boundary between science and religion? It is hard to assess where the borderline is. But it is easy to see when people are far over the line – so far over that it doesn’t really matter where exactly the line is. One can attend any conference on the subject to find people who unabashedly assert that dark matter exists without question. Not just that acceleration discrepancies have been amply demonstrated empirically, but that the only possible interpretation is dark matter. If asked whether this invisible mass is in the room with us now, they will enthusiastically# answer yes! Since dark matter has not been detected in the laboratory, this assertion is an expression of faith – the hallmark of religion – not of an established scientific fact. What we have established is that there are discrepancies between what we see and what we get when we assume Newtonian gravity (or GR, if needed). What we don’t know is whether the cause of these discrepancies is some form of invisible mass (dark matter) or if the equations we employ are inadequate (modified gravity [or more generally, dynamics]).
\n\n\n\nIndeed, these days many people will assert that dark matter has already been detected, usually citing astronomical evidence that used to be considered too feeble to merit a Nobel prize. Funny how repeating a mantra long enough morphs an aspiration into accepted reality. Modern physics is not providing a strong falsification of the supposition that science is a social construct.
\n\n\n\nA prominent example of an observation of the sky that is frequently cited as absolutely requiring cold dark matter is the acoustic power spectrum of the cosmic microwave background. Quoting clayton from a few years ago:
\n\n\n\n\n\n\n\n\nthe primary reason to believe in the phenomenon of cold dark matter is the very high precision with which we measure the CMB power spectrum, especially modes beyond the second acoustic peak. There is a stone-cold, qualitative, crystal clear prediction of CDM about the relative sizes of the second and third peaks that modified gravity profoundly and irredeemably gets wrong: it thinks the third peak should be relatively larger* than the second… whereas CDM thinks they should be about the same
\n
I would accept that this were conclusive proof of dark matter if this were the unique prediction of dark matter: that there was no other way to do it, so all other approaches were indeed irredeemable. (Quite the strong language, eh?) The problem is that CDM is not the one unique was to fit these data. Skordis & Zlosnik showed that it is possible to write a modified gravity theory that also fits the CMB data:
\n\n\n\n![\"\"](\"https://i0.wp.com/tritonstation.com/wp-content/uploads/2023/01/cl_tt_only_skordis.png?resize=700%2C424&ssl=1\")
This does not prove the AeST theory of Skordis & Zlosnik is correct, but it does demonstrate that it is possible to write a modified gravity theory that does indeed do what it is frequently asserted to be impossible for a modified gravity theory to do. I’ve heard of a couple of other theories that can also do this (the relativistic Khronon theory of Blanchet and nonlocal MOND as discussed by Deffayet & Woodard), so clearly this success is not uniquely limited to cold dark matter, or even a particular modified gravity theory. The work of Skordis & Zlosnik (2021) was known and in the literature before clayton made the assertion above in late 2022, so either he wasn’t paying attention (likely) or is convinced that it is impossible so doesn’t even consider the possibility (also likely). The former just says we’re all too busy, but the latter is a mark of religious thinking: my god is the only god, thou shalt have no other hypotheses before& me.
\n\n\n\nMany people are very impressed with the quality of the LCDM fit to the CMB. That is indeed very good, but there are enough free parameters that we were going to get a fit to any physically plausible power spectrum. If not, we’ve never been shy about making up new parameters. (Evolving dark energy, anyone? How about a running power spectrum? There’s a whole bag of possibilities!) What I’ve been more impressed with is the consistency of the fit to the CMB data with the many independent constraints on conventional cosmology. Or at least it was, until it wasn’t.
\n\n\n\nThe Hubble tension has gotten steadily worse (in terms of statistical significance), and it really does not look like local measurements are to blame, nor is it the only tension. People seem to miss that it is the CMB-fitted value of the Hubble constant that has evolved over time to spoil the concordance that got us to believe in LCDM in the first place. But if the CMB is the cornerstone of your religion, all other data must inevitably be at fault and can be ignored: there is an entire community of cosmologists who choose to believe the best-fit Planck cosmology to the exclusion of all other data. It’s like the bad old days of the Hubble tension all over again, with the physics community choosing to believe the lower value of H0 because it makes more sense for the aspects of cosmology that they care about while those in the astronomical community who actually measure H0 find a persistently higher value.
\n\n\n\nA real tension in LCDM implies the need for new physics of the unknown variety. One doesn’t want to go there if it can be helped. I didn’t consider MOND until I was already concerned for the viability of dark matter. There are real problems for the paradigm that its more intense advocates simply deny, brush aside without real thought, or choose to remain ignorant of. When they are confronted with a problem, they are pretty creative about making stuff up on the spot. Anything to avoid having to confront the unspeakable – another hallmark of religion.
\n\n\n\nFor example, cold dark matter is scale free. That’s foundational to the hypothesis. So the existence of an acceleration scale in the kinematic data is anathema to CDM. When I first pointed this contradiction out, there were a variety of assertions to the effect of “does too!” One example is provided by Kaplinghat & Turner, who claim to show “how Milgrom’s law comes about in the cold dark matter theory of structure formation.” That would, indeed, be ideal, and is a requirement for any theory to be successful.
\n\n\n\nWee problem: they demonstrat no such thing. CDM is scale free, yet K&T claim that it explains Milgrom’s Law, which is predicated on the existence of an acceleration scale. Well, which is it? Is CDM scale free? Or does it explains the acceleration scale? We can’t have it both ways: their very premise is self-contradictory. It is absurd on its face.
\n\n\n\nThe acceleration scale is defined by baryons, for which K&T have no model. To connect baryons with dark matter, they make a hand-waving argument about galaxies reaching a0 at the edge of their disks. This is not even a concept of a model and does not begin to suffice as an explanation for many reasons, a prominent one being that low surface brightness galaxies have accelerations less than a0 everywhere:
\n\n\n\n![\"\"](\"https://i0.wp.com/tritonstation.com/wp-content/uploads/2026/04/ARSBeff_forweb.png?resize=700%2C649&ssl=1\")
Milgrom pointed out this and many other shortcomings of their scenario, so I feel no need to elaborate further. Milgrom eviscerated their paper so thoroughly that the proper course of action would have been to retract it. Instead, they simply never acknowledge the criticism, and persist to this day in pushing it as some sort of valid scientific explanation. It is not; it does not withstand even mild critical scrutiny. But it doesn’t need to: it reassures the faithful that all is well. They hear what they want to hear without questioning its veracity. That’s another hallmark of religion.
\n\n\n\nI have refrained from saying these things in the past because I’m too nice. For example, a few years ago I started then abandoned the draft text below, which I simply cut & paste:
\n\n\n\nOne of the things that attracted me to a career in science is the notion of objectivity. I grew up for a time in the bible belt, where people earnestly believed things that were obviously untrue, even to the eyes of a small child. On the occasions that I had the temerity to point out the obvious, the contradictions posed by facts never had an impact on their belief system. Rather, it inevitably earned me a warning that I was going to hell. No few of these people seemed to think it was their religious duty to send me there prematurely, or at least to make life on Earth a living hell.
\n\n\n\nScientists eschew such behavior, but are also human, so often engage in it anyway. I’ve encountered it a lot. I get it; I went through the same denial, grief, and anger over the prospect of losing my good friend cold dark matter. The stages of grief never brought something back from the dead, but it has engendered a lot of blame-the-messenger.
\n\n\n\nHere’s an example, from a review by Mike Turner:
\n\n\n\n![\"\"](\"https://i0.wp.com/tritonstation.com/wp-content/uploads/2022/01/screen-shot-2022-01-16-at-1.35.10-pm.png?resize=700%2C165&ssl=1\")
There is a lot of misinformation packed into this short paragraph.
\n\n\n\nThe first clue is right there at the beginning, in red: the heading “False starts.” This is false framing, a classic tool of propagandists. It starts from the outset by asserting that the topic to be discussed is wrong at a level of knowledge so common it requires no justification. This is not the way one starts an objective discussion, much less a scientific one.
\n\n\n\nTurner then misconstrues what Milgrom did. He didn’t notice the scale a0 in the data, for which there was scant evidence at the time. Rather, Milgrom made the obvious statement that the inference of dark matter relied on the assumption that dynamics, as encapsulated by the laws of inertia and gravity, is the same on the very different scales of galaxies as in the solar system where they were established, so we ought to consider if dynamics might change in some way. He quickly excluded a size dependence as a possibility. How he settled on acceleration is beyond the scope of this post, and not for me to say. Neither is it for Turner to say.
\n\n\n\nAfter a brief and incomplete description of what MOND is, Turner allows that “this one-parameter model fits all the rotation-curve data”. Even in making this admission, he chooses to call it a model rather than a theory. A model is something specific you build in the context of a theory, like a halo model in CDM. MOND is more than that.
\n\n\n\nTurner quickly moves on without contemplating any meaning that rotation curves might hold. Let’s pause to consider that.
\n\n\n\nFirst, I would not say that MOND fits all the rotation curve data. It fits most galaxies, but there are a minority of weird cases that are not well fit. The weird cases inevitably don’t make sense in terms of dark matter either, so on the whole I interpret this to be the usual price of dealing with astronomical data – some of it is just goofy. Setting such cases aside, I can and have fit the same data with all sorts of dark matter halo models. MOND requires fewer parameters, which is important, but the difference isn’t in the fitting. The difference is in predictive ability. I can use MOND to predict the dynamics of galaxies a priori, and have done so many times. I cannot use any flavor of dark matter theory to do the same, and it’s not for lack of trying.
\n\n\n\nThe predictive power of MOND must be telling us something, even if it is something about the nature of dark matter or the process of galaxy formation. There are many papers written on this, some deep and profound, others absurd and banal. Turner cites none of them, nor displays any awareness that such work exists. I would venture to guess that is because acknowledging such work would imply that there is something to debate here, something he would apparently rather not admit.
\n\n\n\nThat’s where I left off. It’s exhausting deciphering other people’s false assertions. Moreover, I just don’t like criticizing other people, no matter how richly they deserve it. (Turner has never refrained from criticizing me in ad hominem terms: on one occasion$ he showed my picture to an audience and called me “the enemy.”) A large segment of the particle physics and cosmology community appears to think this way, and has succumbed to a scientific version of bible thumping in which you can assert any absurd thing so long as it falls within the framework of the holy LCDM. They really need to find something better to do.
\n\n\n\nI had hoped we were past this, but I heard a talk last week that was exactly in this mode. To paraphrase, the talk went
\n\n\n\n\n\n\n\n\nWe’re sure dark matter exists. We have been sure about it for decades. In that time, we have been repeatedly proven wrong about what it is. Rather than re-think our paradigm in the face of these repeated failures, we double down yet again on the existence of this invisible, undetected mass, asserting aggressively% that it must be true while eliding or misrepresenting the evidence that it is not. This enables us to make up a whole lot of exciting new possibilities for what the dark matter might be and conceive of ever more grandiose experiments to continue not to detect it. You must believe in dark matter!
\n
This was not a science talk so much as an indoctrination session. It was as if I had stumbled into a revivalist tent where some hothead was preaching to the choir. This is the kind of talk that misled an entire generation into wasting their careers at the bottom of a mine shaft searching for WIMPs. At least WIMPs were a well-motivated hypothesis; this kind of talk could lead a new generation down an even greater variety of garden paths.
\n\n\n\nI am well aware that I might fall prey to this attitude myself. That’s why I set criteria by which I would change my mind: detect dark matter already, or at least provide a satisfactory explanation as to how MOND comes about. Neither of those criteria have been met. There are claims to do the latter, but so far these are just variations on models I tried and found to fail long ago. If I thought these could work, I would have said so. At the same time, I don’t see any dark matter advocates taking up the challenge to specify what would change their minds. When I ask them what could falsify dark matter, I get dumbfounded looks – the deer-in-the-headlight face one gets when the immediate response why would you even ask that? is checked by a distant memory that scientific theories are supposed to be falsifiable.
\n\n\n\nPersonally, I found it humbling to encounter MOND in my own data. I too thought we understood the universe with dark matter. But who ordered this? Certainly not me: my own conventional, dark-matter based predictions were falsified. No one else working in the context of dark matter had got it right at the time either. Only Milgrom ordered this.
\n\n\n\nAnd what is this? There is a direct connection between what we see and what we get. Even in ignorance of MOND, the radial acceleration relation encodes a one-to-one relation between the distribution of baryons and the effective force. This is so direct that one can right down a single equation connecting the two:
\n\n\n\n\n\n\n\nThe observed acceleration is a simple function of that predicted by Newton for the stars and gas that we see. There is no mention of unseen mass; everything is specified by what we can see is there.
\n\n\n\nI’ve sometimes heard astronomers complain about the reductionist ethos of physics, trying to cram all the complexity of the entire universe into a theory of everything. But here it is appropriate: there is a single, apparently universal force-law at work in galaxies. That’s telling us something profound. And yet if questioned about this, the physicists are the ones who will complain that galaxies are complicated, so they should be exempted from having to explain them. Galaxies should be complicated – in LCDM. But they’re observed not to be, in the sense that a single equation suffices to describe their kinematics. The problem isn’t that galaxies are inexplicably complicated, it’s that they should be but aren’t.
\n\n\n\nI am deeply disappointed that many scientists apparently lack the physical intuition to immediately recognize the import of the simple relation between what we see and what we get. It is the same sort of thing Newton noticed in the solar system: everything happens as if the gravitational force is proportional to the product of the masses and the inverse square of their separation. He didn’t understand why at the time, and was criticized for indulging in magical thinking: how can there be action at a distance? But that’s what the data were saying, and the same applies now. We might not yet understand the why, but that the data look as if MOND is what’s happening in this universe.
\n\n\n\n#The framing has morphed over the years. A recent advent is that some people have started proactively asserting that invisible mass is in the room with us now in order to avoid having to answer it as a question that makes them sound like loonies.
\n\n\n\n*He means the third peak should be smaller than the second, not larger, if by “it” he means modified gravity with the baryon density expected from big bang nucleosynthesis, which was the hypothesis that correctly predicted the first-to-second peak ratio but does indeed get the second-to-third peak ratio wrong. Funny how the CMB community was able to completely ignore the successful prediction for several years, but were then suddenly all over the latter failure. The third peak falsifies the ansatz on which that particular prediction was built, not the entire concept of modified gravity. This would be like asserting that all possible forms of dark matter are excluded because we haven’t yet detected WIMPs. It is a classic failure of objectivity, which is another hallmark of faith-based argumentation: we know His name is [insert favorite deity], not [insert any other deity].
\n\n\n\n&Or after me. Dark matter was my first hypothesis, and I’m here to tell you that True Believers do not suffer second hypotheses or those who stray from the fold. I guess that’s why so many scientists who are MOND-curious keep it on the down low. Wise, perhaps (that’s why tenure needs to be a thing), but hardly the ideal of the open and free exchange of scientific ideas.
\n\n\n\n$I wasn’t there, but one audience member (not someone I knew) thought it was so over the top that he told me about it, sharing a link with a video. (I did not retain that link, and doubt the hosting conference website is still active.)
\n\n\n\n%Argument weak here. RAISE VOICE!
\n","doi":"https://doi.org/10.59350/2rded-9by77","funding_references":null,"guid":"https://tritonstation.com/?p=12487","id":"a6b88cdb-ed8e-48a8-80c6-13fdaa6fda7e","image":"https://i0.wp.com/tritonstation.com/wp-content/uploads/2025/04/nopeaceduckgodrabbitgod.png?fit=1318%2C1470&ssl=1","images":[{"height":"424","src":"https://i0.wp.com/tritonstation.com/wp-content/uploads/2023/01/cl_tt_only_skordis.png?resize=700%2C424&ssl=1","width":"700"},{"height":"649","sizes":"auto, (max-width: 700px) 100vw, 700px","src":"https://i0.wp.com/tritonstation.com/wp-content/uploads/2026/04/ARSBeff_forweb.png?resize=700%2C649&ssl=1","srcset":"https://i0.wp.com/tritonstation.com/wp-content/uploads/2026/04/ARSBeff_forweb.png?resize=1024%2C949&ssl=1, https://i0.wp.com/tritonstation.com/wp-content/uploads/2026/04/ARSBeff_forweb.png?resize=300%2C278&ssl=1, https://i0.wp.com/tritonstation.com/wp-content/uploads/2026/04/ARSBeff_forweb.png?resize=768%2C712&ssl=1, https://i0.wp.com/tritonstation.com/wp-content/uploads/2026/04/ARSBeff_forweb.png?w=1027&ssl=1","width":"700"},{"height":"165","src":"https://i0.wp.com/tritonstation.com/wp-content/uploads/2022/01/screen-shot-2022-01-16-at-1.35.10-pm.png?resize=700%2C165&ssl=1","width":"700"}],"indexed":true,"indexed_at":1777121781,"language":"en","parent_doi":null,"published_at":1777120800,"reference":[],"registered_at":0,"relationships":[],"rid":"h5dzt-szk19","status":"active","summary":"I have been concerned for years that dark matter was morphing from legitimate science into a cold, dark religion. I have been reluctant to put it that way, because there are lots of scientists who work on dark matter that have not fallen entirely down that rabbit hole and who continue to make valuable contributions working in that context.","tags":["Cosmology","Dark Matter","Data Interpretation","LCDM","MOND"],"title":"Yep, it\u2019s a religion","updated_at":1777070683,"url":"https://tritonstation.com/2026/04/25/yep-its-a-religion/","version":"v1"}},{"document":{"abstract":"Co-sponsored by the Union for Democratic Communications, Project Censored and the Park Center for Independent Media Oct.","archive_url":null,"authors":[{"contributor_roles":[],"family":"Akass","given":"Kim"}],"blog":{"archive_collection":null,"archive_host":null,"archive_prefix":null,"archive_timestamps":null,"authors":null,"canonical_url":null,"category":"mediaAndCommunications","community_id":"d0965544-4413-4b89-aedb-36ae2153c1ac","created_at":1730394736,"current_feed_url":null,"description":"Television Studies Blog","doi":null,"doi_as_guid":false,"favicon":"https://rogue-scholar.org/api/communities/d0965544-4413-4b89-aedb-36ae2153c1ac/logo","feed_format":"application/atom+xml","feed_url":"https://cstonline.net/feed/atom/","filter":null,"funding":null,"generator":"WordPress","generator_raw":"WordPress 6.7.1","home_page_url":"https://cstonline.net/","id":"3e29853c-05ee-479f-aa7d-867ff6dce1e9","indexed":true,"issn":null,"language":"en","license":"https://creativecommons.org/licenses/by/4.0/legalcode","mastodon":null,"prefix":"10.59350","registered_at":0,"relative_url":null,"ror":null,"secure":true,"slug":"cstonline","status":"active","subfield":"3315","subfield_validated":null,"title":"CST Online","updated_at":1777191010.224148,"use_api":true,"use_mastodon":false,"user_id":"80307be4-0a5d-4378-a38f-91852e38c1d8"},"blog_name":"CST Online","blog_slug":"cstonline","content_html":"\nggsql: A grammar of graphics for SQL. Posit's alpha release of a SQL extension that adds VISUALIZE, DRAW, PLACE, SCALE, and LABEL clauses to build ggplot-style layered graphics directly from a query result, with a DuckDB backend and support in Quarto, Jupyter, Positron, and VS Code. The stated motivations include meeting analysts who live in SQL on their own ground and the observation that LLMs write SQL fluently, which makes the declarative grammar a natural interface for code-assistant-driven visualization.
![\"\"](\"https://substackcdn.com/image/fetch/$s_!JXra!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd24068cd-34fa-4c30-85f1-5ad83cde0334_682x569.png\")
markdownme: Shameless self promotion here. I published a browser extension that instantly turns any page into markdown with a keyboard shortcut (Alt+M by default). You get an editor window and a preview. You can toggle on/off things like a page map, images, links, page metadata. You can install it from the .xpi file on the GitHub release page, or hopefully soon from the Firefox add-on store (the extension is still under review). If you\u2019re using Chrome (why?) see the credits on the GitHub page and use the upstream maintained by someone else. Conversion on most typical pages takes milliseconds.
![\"\"](\"https://substackcdn.com/image/fetch/$s_!o89D!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4bad732d-fc3e-440b-9c23-238db40048d6_3246x1288.png\")
: Five Things: Mythos and Glasswing, AI medical advice, GPT-Rosalind, bio-AI models and risks.
GPT-5.5 was released yesterday. I haven\u2019t used ChatGPT in a while, but I plan to spend some more time with Codex because I\u2019ve been chewing through my Claude Code session limits in no time lately. Token anxiety is real. And speaking of\u2026
Simon Willison: Is Claude Code going to cost $100/month? Probably not\u2014it's all very confusing. Anthropic silently updated its pricing grid yesterday to remove Claude Code from the $20/month Pro plan and restrict it to Max ($100 and $200 tiers), triggering a several-hour panic across Reddit, HN, and Twitter before reversing the change while Simon was still writing the post. Head of Growth Amol Avasare characterized it as a test on ~2% of new prosumer signups, though Simon (plausibly) thinks that framing doesn't match what people were actually seeing. The piece is less about the specific pricing decision than about how A/B-testing a 5x price increase on a flagship product without any announcement torches trust with the exact audience that teaches and evangelizes the tool; Codex eng lead Thibault Sottiaux capitalized in real time with a promise not to pull the same move. Worth reading as a case study in how much brand damage a canceled experiment can do. I\u2019m a power user of Claude and Claude Code, and now I feel like I\u2019m just waiting for the rug to be pulled.
![\"https://static.simonwillison.net/static/2026/anthropic-x.jpg\"](\"https://substackcdn.com/image/fetch/$s_!C9Ww!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5d2b4a51-39ff-44a2-acc5-8d3dace7e2f4_1444x635.jpeg\" "\\"https://static.simonwillison.net/static/2026/anthropic-x.jpg\\"")
Policy Changes to SBIR and STTR Foreign Disclosure and Risk Management (NOT-OD-26-074). NIH issued guidance implementing the Small Business Innovation and Economic Security Act, signed April 13, 2026, which reauthorizes SBIR/STTR through 2031 and tightens the foreign risk due diligence regime. The notice enumerates categorical denial conditions (malign foreign talent recruitment program participation, any entity, parent, or subsidiary located in the PRC or another country of concern, foreign affiliation of owners or covered individuals with research institutions in those countries, and appearance on any of eight federal risk lists including UFLPA, BIS Entity List, and CMC lists). HHS will not let applicants cure identified security risks before award denial, and material misstatements or ownership changes post-award trigger full repayment.
: Ideas for AI x Bio Hackathon Projects. A practitioner\u2019s wishlist ahead of this weekend\u2019s Apart Research AIxBio Hackathon, from someone who actually operates a DNA screening tool (Commec at IBBIS). Efficient handling of oligo pools and split orders, assembly signature detection, distinguishing abiological DNA (origami, data storage) from biological sequences, better in-silico functionality scoring for predicted SOC variants, and narrowly-scoped customer screening sub-tools like affiliation and address verification APIs.
Annual flu vaccine no longer required for U.S. military. I made the graphic below indicating casualties in the US military during WWI attributed to combat versus those from influenza.
Q&A: Who\u2019s responsible when AI makes mistakes? A short interview with my awesome colleague David Danks, who joined UVA in January with dual appointments in Philosophy and the School of Data Science. David frames the default accountability trajectory as one where humans become perpetual scapegoats: radiologists signing off on AI diagnoses without being given time to actually second-guess them. David argues for product liability as a starting point, floats the more speculative idea of some form of legal personhood for AI systems so they can carry insurance and have seizable property, and pushes back on the race narrative as partly self-generated by companies rather than a reflection of real economic pressure.
Charlie Gao: mori: Shared memory for R objects. First CRAN release of the mori package that places an R object once into OS-level shared memory and lets every worker on the machine read the same physical pages instead of the usual pattern where eight workers with a 1 GB dataset means 8 GB of RAM plus serialization round-trips. Handles atomic vectors, lists, and data frames (so tibbles, data.tables, factors, and matrices come along for free).
![\"\"](\"https://substackcdn.com/image/fetch/$s_!M3wf!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff11f7171-af1e-465c-af9a-50e7c5bdf0fc_1407x491.png\")
: A biosecurity playbook for AI companies. A rundown of five levers frontier labs can pull on biorisk: refusals, misuse classifiers, tiered access with KYC screening, training-time knowledge removal (pretraining data filtration, unlearning, Anthropic's Selective Gradient Masking), and evaluations. Justen pulls useful numbers out of the public record, including Dario Amodei's figure that bioweapons classifiers run around 5% of Anthropic's inference cost and OpenAI's disclosure that safety reasoning can hit 16% during new model rollouts, and draws a sharp analogy between Anthropic's Project Glasswing (vetted-access release of Mythos Preview for cyber defense) and the case for similar gating on biology-capable models.
![\"\"](\"https://substackcdn.com/image/fetch/$s_!7S1k!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fdcffff43-e8bd-4a6d-b493-474c5628fe1a_2048x894.png\")
AI Literacy and Action Lab. UVA Library and the College of Arts & Sciences launched a joint program this year that pairs librarians with faculty as instructional partners across five spring/summer pilots, including Anton Korinek's economics course on AI and the future of work (with hands-on sessions using Claude Code, Google Antigravity, and Codex), Piers Gelly's first-year writing course partnering with a local high school, and early-stage plans with David Danks in the School of Data Science and Andreas Gahlmann in Chemistry. The Lab treats each pilot as a publishable case study with pre/post assessment data, aimed at contributing to a national evidence base for AI literacy in higher education.
![\"\"](\"https://substackcdn.com/image/fetch/$s_!ZraK!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1a382652-c5ea-495a-8273-7a1a2f4a9951_1179x787.png\")
NIH Highlighted Topic: Computational Modeling of Complex Processes Across Biological Scales. A new NIH Highlighted Topic posted April 17, open through April 2027, encouraging investigator-initiated work on computational models that span molecular, cellular, organismal, and epidemiologic scales, with explicit emphasis on replicability, reproducibility, and model reuse. The topic positions multiscale models as a component of Novel Alternative Methods for evaluating mechanism and safety of interventions in preclinical, translational, and clinical development, which aligns with the FDA's ongoing push to reduce animal testing. Highlighted Topics are NIH's lighter-weight priority signal (not a NOFO; applicants go through a Parent Announcement), with ICO funding dependent on availability and meritorious applications. Central contact is Reed Shabman at NIAID.
Purpose: This topic encourages innovative research in computational modeling of complex processes across biological scales (i.e., to develop multiscale models). The topic seeks to build a collaborative community of researchers to improve the replicability and reproducibility of computational multiscale models, promoting their advancement and reuse.
Background: Multiscale computational models that integrate processes across different spatial and temporal levels, from molecular to organismal, to epidemiologic and from microseconds to years. They provide a comprehensive understanding of complex systems and offer an exciting opportunity to advance biomedical research. This approach helps reveal how interactions at molecular and cellular scales influence larger, population-, geographical-, or global-scale phenomena, offering insights into complex biological processes, and may help develop better and more precise biomedical interventions. By integrating processes from molecular to epidemiologic levels, multiscale computational models provide a comprehensive understanding of complex systems. This topic encourages innovative research and collaborative approaches that integrate technologies and informatic practices to develop, improve, and disseminate multiscale computational models for human health and diseases, and their associated technologies, across the research community. The topic also supports leveraging computational multiscale models as an important component of Novel Alternative Methods (NAMs) to investigate the mechanism and safety of a medical intervention in pre-clinical, translational, and clinical development.
Qwen3.6-35B-A3B: Agentic Coding Power, Now Open to All. I started last week\u2019s newsletter with \u201canother week, another new model\u201d where I talked about Claude Opus 4.7 and OpenAI\u2019s GPT-Rosalind. Alibaba's Qwen team just released a new open-weight MoE model with 35B total and 3B active parameters, Apache 2.0 licensed, targeting agentic coding and repo-level reasoning. I hit my Claude token limits while working on a textbook, and switched over to using this local model in Claude Code1 on my 48GB MBP. As much as I\u2019d have loved this to work well, it absolutely did not when compared to just using Sonnet.
Elizabeth Ginexi: The Executive Orders Blocking Your NIH Grant. A 22-year NIH program official walks through the mechanics behind the funding slowdown that followed Congress's FY26 $47 billion appropriation, tracing the pipeline from EO 14168 and the August 7, 2025 \u201cImproving Oversight of Federal Grantmaking\u201d EO through NIH's implementation notices (NOT-OD-25-131, NOT-OD-26-009), the Unified Strategy, and the elimination of paylines as funding cutoffs. The piece discusses the requirement that political appointees personally approve NOFOs and discretionary awards, the termination clause that now applies to renewals and continuations, and the provision that court victories only bind NIH for the specific plaintiffs who sued. Ginexi also documents the collapse in new NOFOs (756 in 2024, 17 through mid-March 2026) and the shift from direct indirect-cost caps (struck down in court) to preference-based screening at the appointee review stage.
And ICYMI, I published several posts here this week.
New papers & preprints:
You can run Claude Code with local models via Ollama. E.g.: ollama launch claude --model qwen3.6
by Olesia Zvezdova
\nIs it possible to change the past? Not in the sense of speculative fiction, nor as an attempt to correct individual mistakes, but as a strategy aimed at securing advantage in large-scale geopolitical competition. To what extent can collective memory of historical events be instrumentalized as a weapon against political adversaries? Is it possible to rewrite the pages of history textbooks and shape a new generation that will sincerely believe in a newly constructed truth? These questions have repeatedly arisen in my research on the key mechanisms of psychological influence employed by the Russian Federation, both with regard to its domestic population and to audiences in European countries. Among the dominant narratives disseminated through Russian and pro-Russian media channels, the manipulation of historical events and historical figures occupies a central place. Therefore, during my fellowship at the IEG DH Lab, I focused my research on historical revisionism in social media as a method of Russian psychological aggression.
\n
We can consider psychological warfare to be a form of information warfare, given that natural and artificially created information influences a person’s mental state, as evidenced by their thoughts, actions and behaviour (Renz 2016). Just as a person navigates the external world, they also navigate their internal world of mental phenomena. However, this inner world is often altered by external informational factors. Attitudes towards historical facts and events can shape stereotypes and influence the behaviour of population groups, thereby also affecting perceptions of contemporary situations.
\nAnyone who actively uses social media will be familiar with the Russian Federation\u2019s dominant historical narratives, such as \u201cUkraine was created by Lenin\u201d, \u201cUkrainians and Russians are the same nation\u201d, \u201cUkraine has no history without Russia.\u201d In addition, nostalgic references to the Soviet Union are frequently mobilised, and the ongoing war is often framed through analogies with the \u201cGreat Patriotic War,\u201d thereby invoking emotionally charged historical memory to legitimise contemporary aggression. Widely disseminated across multiple digital platforms, these claims represent a form of historical revisionism mobilised to shape and influence public perception.
\nThe most important channel for disseminating this information is social media, as it enables the Russian state and its affiliates to reach a wide range of population groups. Through platforms such as Facebook, Telegram, YouTube, and Twitter, historical narratives, revisionist claims, and emotionally charged messaging can be rapidly circulated to maximise influence and impact. Social media thus serves as a key tool in Russia\u2019s psychological operations. Given the psychological impact of historical narratives, it is crucial to assess the role of historical revisionism in Russian psychological operations on social media.
\nHistorical revisionism is considered as a contested and multifaceted phenomenon in modern political science. The positive effects and even the necessity of historical revisionism are often emphasized by different scholars. At the beginning of the twentieth century, Robert Clark emphasised that history should be revised in order to correct the mistakes of previous generations of researchers following the discovery of new facts (Clark 1932). Jason Morgan argues that historical revisionism is an inherent part of modern scholarship, as much historical writing is produced by members of dominant social groups. He considers historical revisionism as \u201c… a rewriting of history to bring into historiographical focus, either better or perhaps for the first time, past traumas that have been insufficiently examined in historical records and writing.\u201d (Morgan 2025).
\nFrom another perspective, historical revisionism can undermine the integrity of historical understanding and contribute to societal division and disinformation. Such historical disinformation is often created and disseminated intentionally, distorting or misinterpreting past events in order to serve various agendas, including political, ideological, or economic objectives (Singh and Kumar 2024). Antoon De Baets proposes the concept of \u201chistorical abuse\u201d in such cases and emphasises three levels of this phenomenon:
\nIt reflects the core features of Russian revisionism of Ukrainian history, namely the suppression of unwelcome narratives (genocide denial) in combination with imperial ambitions (commissioned histories). The advent of digital technologies has revolutionised the ways in which information is produced, shared, and consumed (Ejaz and Felix 2025).
\nAmong the most consequential means of information dissemination are social media platforms, which not only enable outreach to broad segments of the population with targeted historical narratives but also facilitate the distortion of public perception and the reframing of historical events to serve specific ideological or political agendas. And it becomes increasingly crucial to verify sources to prevent the dissemination of false narratives.
\nWhile historical revisionism is produced and legitimized within official political and educational frameworks in Russia (Putin 2021), social media serve as the primary channel through which these narratives are disseminated to and internalized by the broader population. Accordingly, the primary gap in the existing literature lies in the insufficient analysis of Russian historical revisionism in social media and its impact on the formation of public opinion.
\nSocial media platforms are strategic communication spaces where historical narratives can be amplified, contested, or weaponized to influence public opinion. The platforms could be classified according to their primary mode of content creation and interaction. A widely used framework distinguishes platforms into three categories (Table 1):
\n| Category | \nDefinition | \nExamples | \n
| Content-sharing platforms | \nplatforms where the focus is on sharing multimedia content such as videos, images, or music. Users primarily consume and distribute content rather than engage in extended textual interactions. | \nYouTube, RuTube | \n
| Networking platforms | \nplatforms primarily designed for social connections and community building, emphasizing user interaction, discussion, and sharing of diverse content types (text, images, videos, links). | \nFacebook, VKontakte | \n
| Micro-blogging platforms | \nplatforms that focus on short, real-time textual updates and rapid information dissemination, often emphasizing speed over depth of content. | \nX/Twitter, Telegram | \n
I focus on the role of YouTube and Facebook due to their widespread use, diverse content formats, and significant influence on public discourse. Russian domestic platforms, such as RuTube and VKontakte, were excluded because they are inaccessible in Ukraine, while microblogging platforms were excluded due to their focus on short-form content, which lacks the depth required for analyzing historical narratives. By concentrating on YouTube and Facebook, the study captures both video-centric content dissemination and network-based discussion and engagement, enabling a comparative analysis of how historical revisionist narratives are produced, circulated, and received in different digital environments.
\nMy study employs a qualitative-quantitative mixed-methods approach to examine the dissemination of Russian historical revisionism in social media and its influence on public opinion. The methodological framework combines content analysis and elements of digital ethnography in order to capture both the production of revisionist narratives and the mechanisms of their circulation and reception in online environments.
First, official texts produced by Russian political actors, including policy documents, state-approved educational materials, and public speeches, produce dominant historical narratives in line with contemporary political objectives. Qualitative and quantitative content analysis is used to examine social media materials that reproduce, adapt, or amplify these revisionist narratives. The analysis focuses on posts, visual materials, and hashtags disseminated through major platforms. Quantitative indicators (frequency, engagement metrics, repost networks) are combined with qualitative coding to assess narrative persistence, emotional framing, and patterns of normalization.
\nSecond, elements of digital ethnography are employed to contextualize how users interact with revisionist content. This includes observing comment sections and discussion threads in order to understand how historical narratives are interpreted, contested, or internalized by different audiences. Finally, a comparative approach is employed to examine and contrast the dominant historical narratives disseminated on different platforms, with attention to platform specific affordances and modes of audience engagement.
\nI focus on the following domains where Russian psychological operations employ historical revisionism, each with its own set of narratives.
\nEducational domain \u2013 the dissemination of historical disinformation that distorts historical events and portrays Ukrainian historical leaders in a negative light. It is aimed at constructing a distorted understanding of the past not only among the Ukrainian population but also within the broader European audience. The introduction of historical disinformation into educational channels enables broad audience reach, particularly among individuals with an interest in historical content. The dominant narratives within this domain focus on the activities of the Organization of Ukrainian Nationalists (OUN) and the Ukrainian Insurgent Army (UPA) during the Second World War, the figure of Stepan Bandera, and the portrayal of Ukraine as having an inherently Nazi past.
\nSocial domain \u2013 the propagation of narratives asserting that the Ukrainian population has historically constituted an integral part of the Russian people, thereby undermining Ukrainian national identity. The dominant narratives focus on the issue of the establishment of Ukrainian statehood, which is artificially presented as being exclusively dependent on the political will of Russian political elites (mainly, Vladimir Lenin and Josif Stalin). These narratives target pupils and students, who represent a particularly vulnerable audience due to their formative stage of identity and historical understanding.
\nMilitary domain \u2013 the systematic circulation of fabricated or exaggerated reports of military success intended to reinforce the perception that resistance against Russia is futile. The symbolic instrumentalization of \u201cVictory Day\u201d in the Second World War functions as a recurring marker of Russian military power and historical inevitability (Juurvee et al. 2021). The image of Great Patriotic War is used not only to emphasize the narrative of sacrifice associated with the Russian army but also to underscore a purported common historical past shared by Russia and Ukraine.
\nInterpersonal communication domain \u2013 the use of so-called informational \u201cintrusions\u201d within social networks. Historical revisionist narratives (such as denying Holodomor 1932-1933, Stalins repression, or nostalgia for common Soviet past) designed to activate contested historical memories. These intrusions increase psychological polarization, contribute to social fragmentation rather than social cohesion.
\nThis multi-domain strategy demonstrates how historical narratives can be instrumentalized as tools of psychological operations and information warfare, with significant implications for both domestic and international audiences.
\nTo end this brief report, psychological warfare constitutes an integral component of Russia\u2019s broader strategy in areas where it perceives its national interests to be at stake, with historical revisionism serving as a key instrument. Within this framework, historical revisionism functions as a manipulative tool that supports authoritarian governance and advances the Russian Federation\u2019s revisionist foreign policy objectives (Arribas et al. 2023). It is employed to undermine societal resilience, deny Ukraine\u2019s national identity, and legitimise the war of aggression.
\nFeatured Image: Illustration of the historical evolution
\nof Russian psychological influence in mass media (top: artwork by Y. Scheglov; bottom: author\u2019s screenshot from YouTube)
OpenCitations has formally adopted the Principles of Open Scholarly Infrastructure (POSI) since its first self-assessment in August 2021. At that time, we had only recently been included in the SCOSS funding round and, although we had a clear vision of what we wanted to build and the role we wanted to play within the Open Science ecosystem, both our financial and human resources were still very limited. For this reason, the POSI self-assessment proved to be an important exercise, since it allowed us to critically reflect on both the strengths we could build upon and the weaknesses we needed to address as we developed OpenCitations into a sustainable, community-governed Open Science infrastructure.\u00a0
\nThat first assessment highlighted some key areas of attention. In particular, it highlighted the limited executive power of the community bodies involved in governance, as well as challenges related to long-term financial sustainability, particularly the ability to generate and manage financial surpluses.\u00a0
\nBetween 2021 and 2026, OpenCitations has evolved significantly, becoming a more complex infrastructure by expanding both the services it offers and the number of data sources it integrates. During this period, we have also expanded our team, worked on the development of a container-based infrastructure, and relaunched our governance framework.\u00a0
\nThese developments made it necessary to update our POSI self-assessment. This update is also timely, given the recent release of POSI Version 2 in October 2025. The new version of the principles is the result of a collective effort by POSI adopters and includes several clarifications to existing principles (for example, around the concept of lobbying and the distinction between transparent governance and transparent operations). In addition, new principles have been introduced to better reflect how scholarly infrastructures operate within the broader research ecosystem.\u00a0
\nWe are therefore happy to share our updated POSI self-assessment, which provides a comprehensive snapshot of OpenCitations as of April 2026. At the same time, it represents a renewed declaration of our commitment to the POSI principles.\u00a0
\nDisclaimer. As with many early POSI adopters, our previous assessment used a traffic-light system to indicate the level of alignment with each principle. In this system:\u00a0
\nDuring the POSI Version 2 working group discussions, it emerged that the use of this traffic-light system is not mandatory for POSI self-assessments. Nevertheless, we have decided to maintain this approach, as we believe it remains particularly effective as a communication tool and, at the same time, ensures continuity with our previous assessment. We have, however, taken the liberty of adding one more element: a symbol to indicate where we were already \u201cgreen\u201d in 2021, but have further improved our performance.\u00a0
\nThe symbol is therefore a \u201cplus\u201d placed next to the green traffic light.
\n![\"\"](\"https://opencitations.hypotheses.org/files/2026/04/Green-Simple-Project-Comparison-Chart-Graphic-3-e1776933564657.png\")
Coverage across the scholarly enterprise
\nOpenCitations demonstrates broad coverage by collecting citation data from global scholarship, ensuring representation across disciplines, geographic areas, and research communities. The scope of OpenCitations’ coverage is universal, not limited to a particular scholarly domain, nor to the English language, nor restricted by imposed acceptance criteria.
\nStakeholder governed
\nGovernance is structured to reflect the stakeholder community, with the International Advisory Board elected by the Council of members and responsible for approving the Trustee Network. For more information: https://opencitations.net/governance/\u00a0
\nNon-discriminatory participation or membership
\nMembership is open to all individuals and organisations that support Open Science principles, enabling inclusive and non-discriminatory participation in line with OpenCitations\u2019 founding values.
\nTransparent governance
\nOpenCitations maintains a high level of transparency by publicly documenting its organisational structure, governance processes, and decision-making procedures. Reports and relevant documentation are shared with the community, including annual reports and the Rules of Membership and Organizational Bodies.
\nCannot lobby
\nOpenCitations does not engage in political or financial lobbying activities. Its role remains focused on supporting the scholarly community without pursuing regulatory changes that would advantage its own position.
\nLiving will
\nA clear long-term stewardship strategy is in place, ensuring that data, services, and infrastructure can be responsibly transferred if needed. This is supported by a recently implemented fully replicable technical infrastructure and a new governance model designed to facilitate handover thanks to the presence of the Trustee Network, which is capable of appointing new hosting members to physically and administratively host the infrastructure.
\nRegular review of purpose and community value
\nOpenCitations has recently monitored its relevance and community value through mechanisms such as community surveys. In addition, it has established governance and technical strategies to support a responsible wind-down process, including, if necessary, the transfer of assets and infrastructure through its Trustee Network. Additionally, the Trustee Network is responsible for regularly monitoring OpenCitations\u2019 adherence to its mission and values, as well as its annual activities and finances.
\n![\"\"](\"https://opencitations.hypotheses.org/files/2026/04/Green-Simple-Project-Comparison-Chart-Graphic-4-e1776933649286.png\")
Transparent operations
\nOpenCitations ensures a high level of operational transparency by openly providing key documentation, including financial reports, a public roadmap, a mission statement and value proposition, a sustainability model and fee structure, and privacy policies.
\nTime-limited funds used only for time-limited activities
\nGrant income is restricted to funding specific, time-limited projects, including the appointment of personnel working on them, while core operations are not dependent on such funding.\u00a0
\nGoal to generate surplus
\nThanks to memberships and donations, as well as in-kind support from the University of Bologna, OpenCitations has recently achieved the financial capacity to ensure stability until 2029, at least, in terms of funds allocated for staff salaries and technical operational expenses.
\nEstablish and maintain financial reserves guided by policy
\nAlthough OpenCitations has reached a level of financial stability that ensures a budget surplus, there is currently no formal financial policy defining the amount of reserves to be allocated for a transition or wind-down plan, or to address exceptional or unforeseen events. However, OpenCitations has already initiated the necessary consultations to develop a Financial Reserve Policy, which will not only define the level and management of reserves but also provide a clear framework for handling revenues. In addition, it will establish and formalise procedures for approving both budget forecasts and actual expenditures.
\nMission-consistent revenue generation
\nRevenue generation is aligned with the organisation\u2019s mission (in particular, the value proposition according to which \u201cexternal financial support is required from the stakeholder community to support OpenCitations and enable it to expand its delivery of high-quality comprehensive open bibliographic and citation metadata\u201d), primarily through community funding via membership fees and donations. OpenCitations\u2018 members are listed on the website: https://opencitations.net/members-and-donors/\u00a0\u00a0
\nRevenue generated from services, not data
\nOpenCitations charges no fees for any of its services, access to its data, or reuse of its software. OpenCitations members, donors and third parties all have equal free access.
\nVolunteer labour
\nOpenCitations\u2019 core operations are carried out by paid staff, ensuring that the continuity and reliability of its services do not depend on volunteer labour. Nevertheless, OpenCitations recognises the value of voluntary labour. Indeed, the document Rules of Membership and Organizational Structure specifies that membership of the International Advisory Board is honorary and without remuneration, and that the Hosting Entity will reimburse all reasonable expenses related to travel, accommodation, and meals incurred in attending meetings, within the limits of the budget allocated for such purposes. While this is already stated in the governance framework, it is important to reiterate and further formalise it within the Financial Policy document currently under development. More broadly, the OpenCitations Mission Statement emphasises the importance of community engagement (voluntary and non-remunerated) through, for example, the involvement of community actors in the direct provision and curation of OpenCitations data, as well as the broader role of the community in supporting funding and participating in governance.
\nTransition planning
\nTransition planning is only partially developed. While the governance structure is in place, ensuring a management handover through the possibility of changing the hosting member upon approval by the Trustee Network, there is a lack of detailed operational documentation for individual roles within the management and technical teams, which may limit the organisation\u2019s ability to ensure immediate continuity in the event of key personnel changes.
\n![\"\"](\"https://opencitations.hypotheses.org/files/2026/04/Green-Simple-Project-Comparison-Chart-Graphic-5-e1776933754542.png\")
Open source
\nAll OpenCitations software is released under open source licences, ensuring full transparency, reusability, and the possibility for the community to inspect, modify, and replicate the infrastructure.
\nOpen data
\nTo ensure the greatest possible reusability, all OpenCitations data is published under a Creative Commons CC0 Public Domain Waiver that permits downloading and re-use of any nature, including added-value re-purposing and commercial exploitation.
\nAvailable data
\nData are provided through multiple access points, including REST APIs, SPARQL endpoints, query interfaces, and downloadable data dumps. This ensures broad accessibility and supports diverse use cases across the community.
\nPatent non-assertion
\nOpenCitations commits not to pursue patents, ensuring its infrastructure remains fully open and replicable.\u00a0
\nPrioritise interoperability and open standards to ensure continuity and resilience
\nThe infrastructure has been redesigned as a container-based solution, thereby facilitating replication, deployment across different environments, and long-term service continuity.
\n","doi":"https://doi.org/10.59350/22h4w-czb47","funding_references":null,"guid":"https://opencitations.hypotheses.org/?p=4260","id":"4a7d7213-1f03-4891-a399-2ce6fb6d8173","image":"https://opencitations.hypotheses.org/files/2026/04/POSI-v2.0.png","images":[{"height":"768","sizes":"auto, (max-width: 990px) 100vw, 990px","src":"https://opencitations.hypotheses.org/files/2026/04/Green-Simple-Project-Comparison-Chart-Graphic-3-e1776933564657.png","srcset":"https://opencitations.hypotheses.org/files/2026/04/Green-Simple-Project-Comparison-Chart-Graphic-3-e1776933564657.png, https://opencitations.hypotheses.org/files/2026/04/Green-Simple-Project-Comparison-Chart-Graphic-3-e1776933564657-300x233.png, https://opencitations.hypotheses.org/files/2026/04/Green-Simple-Project-Comparison-Chart-Graphic-3-e1776933564657-500x388.png, https://opencitations.hypotheses.org/files/2026/04/Green-Simple-Project-Comparison-Chart-Graphic-3-e1776933564657-768x596.png","width":"990"},{"height":"768","sizes":"auto, (max-width: 957px) 100vw, 957px","src":"https://opencitations.hypotheses.org/files/2026/04/Green-Simple-Project-Comparison-Chart-Graphic-4-e1776933649286.png","srcset":"https://opencitations.hypotheses.org/files/2026/04/Green-Simple-Project-Comparison-Chart-Graphic-4-e1776933649286.png, https://opencitations.hypotheses.org/files/2026/04/Green-Simple-Project-Comparison-Chart-Graphic-4-e1776933649286-300x241.png, https://opencitations.hypotheses.org/files/2026/04/Green-Simple-Project-Comparison-Chart-Graphic-4-e1776933649286-500x401.png, https://opencitations.hypotheses.org/files/2026/04/Green-Simple-Project-Comparison-Chart-Graphic-4-e1776933649286-768x616.png","width":"957"},{"height":"730","sizes":"auto, (max-width: 975px) 100vw, 975px","src":"https://opencitations.hypotheses.org/files/2026/04/Green-Simple-Project-Comparison-Chart-Graphic-5-e1776933754542.png","srcset":"https://opencitations.hypotheses.org/files/2026/04/Green-Simple-Project-Comparison-Chart-Graphic-5-e1776933754542.png, https://opencitations.hypotheses.org/files/2026/04/Green-Simple-Project-Comparison-Chart-Graphic-5-e1776933754542-300x225.png, https://opencitations.hypotheses.org/files/2026/04/Green-Simple-Project-Comparison-Chart-Graphic-5-e1776933754542-500x374.png, https://opencitations.hypotheses.org/files/2026/04/Green-Simple-Project-Comparison-Chart-Graphic-5-e1776933754542-768x575.png","width":"975"}],"indexed":true,"indexed_at":1777032089,"language":"en","parent_doi":null,"published_at":1777016681,"reference":[],"registered_at":0,"relationships":[],"rid":"9p94s-9tf12","status":"active","summary":"OpenCitations has formally adopted the Principles of Open Scholarly Infrastructure (POSI) since its first self-assessment in August 2021. At that time, we had only recently been included in the SCOSS funding round and, although we had a clear vision of what we wanted to build and the role we wanted to play within the Open Science ecosystem, both our financial and human resources were still very limited.","tags":["Governance And Finances","Non Class\u00e9","Open Science","OpenCitations","Governance"],"title":"OpenCitations\u2019 renewed compliance with the Principles of Open Scholarly Infrastructure (April 2026)","updated_at":1777018363,"url":"https://opencitations.hypotheses.org/4260","version":"v1"}},{"document":{"abstract":"En la entrada anterior discut\u00ed el origen de la falsa autoridad b\u00edblica Tres vidit et unum adoravit y su uso en el Lucidario y en los Castigos de Sancho IV. La primera investigadora en notar esta presencia fue Francis Gormly, en una tesis doctoral dedicada al uso de la Biblia en la literatura...","archive_url":null,"authors":[{"contributor_roles":[],"family":"Coss\u00edo Olavide","given":"Mario","url":"https://orcid.org/0000-0002-1447-3891"}],"blog":{"archive_collection":null,"archive_host":null,"archive_prefix":null,"archive_timestamps":null,"authors":null,"canonical_url":null,"category":"languagesAndLiterature","community_id":"66270be2-3364-4cdc-8884-bad0e29f524f","created_at":1705393153,"current_feed_url":null,"description":"Editando el \u00abLucidario\u00bb de Sancho IV","doi":null,"doi_as_guid":false,"favicon":"https://rogue-scholar.org/api/communities/66270be2-3364-4cdc-8884-bad0e29f524f/logo","feed_format":"application/atom+xml","feed_url":"https://lucidarios.hypotheses.org/feed/atom/","filter":null,"funding":null,"generator":"WordPress","generator_raw":"WordPress 6.6.2","home_page_url":"https://lucidarios.hypotheses.org","id":"9e41cc54-c38e-495d-b65c-38cf327d3c73","indexed":true,"issn":"2952-1580","language":"es","license":"https://creativecommons.org/licenses/by/4.0/legalcode","mastodon":null,"prefix":null,"registered_at":1708118497,"relative_url":null,"ror":null,"secure":true,"slug":"lucidarios","status":"active","subfield":"3310","subfield_validated":null,"title":"Lucidarios","updated_at":1777191776.28654,"use_api":true,"use_mastodon":false,"user_id":"4ef2e0ac-b1f9-433b-92b8-7df17cc1feb7"},"blog_name":"Lucidarios","blog_slug":"lucidarios","content_html":"\nEn la entrada anterior discut\u00ed el origen de la falsa autoridad b\u00edblica Tres vidit et unum adoravit y su uso en el Lucidario y en los Castigos de Sancho IV. La primera investigadora en notar esta presencia fue Francis Gormly, en una tesis doctoral dedicada al uso de la Biblia en la literatura espa\u00f1ola medieval. Gormly se\u00f1ala que los predicadores medievales citaban a menudo la liturgia en lugar del texto de las Escrituras, porque la Biblia muchas veces carec\u00eda de declaraciones expl\u00edcitas sobre ciertos dogmas centrales a la fe (72), como el dogma trinitario.
Gormly afirma esto a partir de su lectura del Artes praedicandi de Thomas-Marie Charland, un volumen dedicado a las artes praedicandi publicado en la primera mitad del siglo 20, en el que tambi\u00e9n se editan dos importantes tratados de predicaci\u00f3n del siglo 14. En su estudio, Charland nota que esta falsa autoridad fue r\u00e1pidamente incorporada a la liturgia y a menudo empleada por los predicadores medievales en sus sermones (114), como denuncian las dos artes praedicandi que edita.
En la primera de ellas, la Forma praedicandi (1322) de Roberto de Basevorn, este predicador de origen incierto relata haber escuchado el pasaje empleado como thema en un serm\u00f3n del Domingo de la Trinidad, a pesar de no proceder de un vers\u00edculo b\u00edblico. Como recuerda Bizzarri, \u201cun serm\u00f3n no es otra cosa que una reflexi\u00f3n sobre alg\u00fan pasaje b\u00edblico\u2026 que se eleg\u00eda como reflexi\u00f3n del d\u00eda\u201d (206\u201307). En la Forma praedicandi:
Videat etiam quod thema suum sit de textu Bibliae, non de antiphonario. Unde vitiosum est illud thema quod aliqui assumunt in festo Trinitatis: Tres vidit et unum adoravit, quia non est textus Biblia (Charland 250).
M\u00e1s extensa es la discusi\u00f3n de Thomas Waleys, te\u00f3logo dominico y profesor de teolog\u00eda en la Universidad de Oxford en el siglo 14, quien, en De modo componendi sermones cum documentis (mediados del siglo 14), relata un episodio similar. Waleys comienza explicando que muchas veces los c\u00e1nticos sacros se inspiran en la Escritura, pero que existe una gran distancia entre el texto b\u00edblico y las canciones religiosas (contigit enim quandoque quod ea quae cantantur in Ecclesia de Sacra Scriptura extrahuntur, sed inter originale et transsumptum est magna diversitas, 342). En ocasiones, afirma Waleys, incluso un excelente predicador puede errar al usar como fuente los antifonarios y no la Biblia. Como Roberto, Waleys relata haber escuchado a un predicador utilizar esta cita \u2014que \u00e9l identifica como originada en el primer nocturno del Domingo de Quincuag\u00e9sima\u2014 como thema para un serm\u00f3n:
Praedicaturus de Sancta Trinitate, accepit istud thema: Tres vidit et unum adoravit. Quia enim illud continetur in quodam responsorio Dominicae in quinquagesima, quod responsorium transsumptum est de Genesi, capitulo 18, credebat quod illa verba continerentur ibidem, cum tamen nullibi reperiantur in Sacra Scriptura.
Tres a\u00f1os despu\u00e9s del trabajo de Gormly, Catal\u00e1n public\u00f3 un art\u00edculo-resumen sobre esta tesis doctoral, en el que a\u00f1adi\u00f3 otro ejemplo de la presencia de esta falsa autoridad en la Fazienda de Ultramar. En un relativamente reciente estudio sobre la Biblia en los sermones castellanos durante la Edad Media, Bizzarri apunta que el comportamiento detectado por Charland, Gormly, Catal\u00e1n (y antes, Roberto y Thomas) responde a una tendencia: los predicadores comenzaron a alejarse de la Biblia y a recurrir a otros textos, como falsas autoridades y refranes, para construir sermones m\u00e1s convincentes e incitar a los fieles a la devoci\u00f3n (207).
\n\n\n\nEste es el mismo patr\u00f3n que encontramos en las dos citas b\u00edblicas del Lucidario mencionadas en las entradas anteriores, lo que puede apuntar a que en la redacci\u00f3n del texto participaran predicadores del entorno cortesano de Sancho IV. Una evidencia adicional puede encontrarse en algo que Gormly subraya sobre dos formas de citar en los Castigos, tambi\u00e9n presentes en el Lucidario, tambi\u00e9n propias de los predicadores medievales: la falsa atribuci\u00f3n de aforismos al rey Salom\u00f3n (por ejemplo, en el cap\u00edtulo 22: \u2026dixo el rey Salam\u00f3n: \u201cEx abondantia cordis, os loquitur\u201d, cita tomada de un pasaje que aparece en Mt\u00a012.34 y Lc\u00a06.45) y la identificaci\u00f3n incorrecta de textos veterotestamentarios como pasajes evang\u00e9licos (73).
Obras citadas
\n\n\n\nBizzarri, Hugo. \u201cLa Biblia en la prosa homil\u00e9tica y moral de la Edad Media.\u201d La Biblia en la literatura espa\u00f1ola. Vol. 1.2: La Edad Media. El texto: fuente y autoridad, editado por Mar\u00eda Isabel Toro Pascua. Trotta, 2008. 195\u2013252.
\n\n\n\nCatal\u00e1n, Diego. \u201cLa Biblia en la literatura medieval espa\u00f1ola.\u201d Hispanic Review, vol. 33, no. 3, 1965, pp. 310\u201318. https://doi.org/10.2307/471842.
\n\n\n\nCharland, Thomas-Marie, O. P. Artes praedicandi. Contribution \u00e0 l\u2019histoire de la rh\u00e9torique au Moyen \u00c2ge. Librairie Philosophique J. Vrin, 1936.
\n\n\n\nGormly, Francis. The Use of the Bible in Representative Works of Medieval Spanish Literature, 1250\u20131300. The Catholic University of America, 1962. Tesis doctoral.
\n\n\n\nImagen de miniatura: Bible historiale,
New York, The Morgan Library & Museum, ms. M.394, fol. 17v.
![\"\"](\"https://i0.wp.com/blogatarraya.com/wp-content/uploads/2026/04/16-Chimal-Prohibicion.png?resize=770%2C770&ssl=1\")
Regulaci\u00f3n de la venta y consumo de pulque en Nueva Espa\u00f1a: de su prohibici\u00f3n a su legalizaci\u00f3n por razones fiscales.
\n\n\n\nAbraham Chimal, \u00abLa regulaci\u00f3n de bebidas embriagantes en M\u00e9xico/Nueva Espa\u00f1a. Un marco amplio para la reflexi\u00f3n sobre esquemas de prohibici\u00f3n, Iberoamericana, vol. 22, n\u00fam. 83, 2023, pp. 83-106. https://doi.org/10.18441/ibam.23.2023…
\n\n\n\nTambi\u00e9n disponible en otras plataformas
\n\n\n\n![\"\"](\"https://i0.wp.com/blogatarraya.com/wp-content/uploads/2026/04/Captura-de-pantalla-2026-04-23-a-las-8.49.41-a.m.png?resize=770%2C391&ssl=1\")
20:20 hindsight on scientific progress
Open research in the humanities
The role of scientific sleuths
Individual bias and collective criticism
Whitney Barlow Robles looks back over the history of scientific progress and notes that a lot of it would look \u201cunscientific\u201d by today\u2019s standards . . . which will look \u201cunscientific\u201d in the future!
People sometimes look back at past scientists and ask:
\u201cCould they really have been this dumb? Thank goodness we know better.\u201d
In response, Robles argues that science sometimes \u201cprogresses\u201d in terrible ways, and that:
Brushing off these practices as unscientific implies that \u201ctrue\u201d science is, and always has been, unimpeachable. It denies us a chance to examine what flaws may fuel our own ways of thinking. And it absolves us from reckoning with the wreckage of the scientific past.
Robles, W. B. (2026, April 16). They thought what? Contingent Magazine. https://contingentmagazine.org/2026/04/16/they-thought-what/
Beatriz Ferreira\u2019s recent article considers open research in the humanities:
[It] contributes to the broader discussion on the need to reframe openness not as a set of procedural norms or policy mandates, but as a relational and epistemically situated practice.
Ferreira, B. B. (2026). Open for debate: Situating open research for the humanities in a neoliberal setting. The Journal of Electronic Publishing 29(1). https://doi.org/10.3998/jep.7850
This article is part of a special issue on \u201cOpen Research for the Humanities and Social Sciences,\u201d edited by Jenni Adams, , and Samuel Moore: https://journals.publishing.umich.edu/jep/issue/456/info/
Serge Horbach and Auste Valinciute interrogate the emerging and contested role of the \u201cscientific sleuth\u201d in forensic metascience.
[Sleuths] form a distinct class of actors, most importantly because of the absence of formal accountability structures. This has triggered comparisons of scientific sleuthing to vigilantism.
Serge, P. M. J., & Horbach, A. V. (2026). Contextualizing the emerging and contested role of sleuths in the science monitoring system. Science and Public Policy, scag044, https://doi.org/10.1093/scipol/scag044
Commenting on a recent paper by Shiffrin et al. (2026), Steven Sloman and Almos Molnar argue that:
The target article provides an abundant collection of biases that influence scientists .... But all of this depends on viewing scientific activity according to what Chater and Loewenstein (2023) refer to as an 'i-frame,' that is, a focus on the behavior of individuals.
In contrast, they consider the collective adversarial dynamics of the scientific community and conclude that:
In this collective dynamic, there is even a \u2018methodological justification for individual scientists to be biased\u2019, because only when ideas are tenaciously defended by some and tenaciously criticized by others, can we discern their true merit (Popper, 1994: 93).
Sloman, S. A., & Molnar, A. C. (2026). Are scientists na\u00efve humans? Computational Brain and Behavior. https://doi.org/10.1007/s42113-026-00296-6
The following is an edited transcript of oral remarks delivered by Reda Sadki on 26 April 2026, on the occasion of the fifth anniversary of Teach to Reach. The text has been lightly tidied for clarity, grammar, and flow.
\n\n\n\nTeach to Reach is a peer learning platform, network, and community hosted by The Geneva Learning Foundation TGLF). Launched in January 2021, it has grown into a multilingual gathering where frontline health and humanitarian workers, policymakers, researchers, and global partners meet, network, and learn together around urgent challenges such as immunization, climate change and health, malaria, neglected tropical diseases, gender equity, and humanitarian response. Teach to Reach is one of three complementary pedagogical patterns developed by Reda Sadki, TGLF\u2019s founder, alongside the peer learning exercise and the Impact Accelerator.
\n\n\n\nTogether, these models operationalize Sadki’s learning theory, which posits that active knowledge production, structured emergence, networked accountability, and scale as a driver of quality form a complete learning ecosystem. On this fifth anniversary, Reda Sadki traces the origins of Teach to Reach, describes how it works before, during, and after the live event, and situates it within the Foundation’s broader system for turning peer learning into measurable impact on health outcomes.
\n\n\n\nMy focus today is on Teach to Reach, although there is, of course, the broader history of the Geneva Learning Foundation behind it.
\n\n\n\nWhat we have been chasing after for a long time is really this question: in the digital age, and now in the age of AI, how do we build meaningful connections that help us better understand what it means to be human, and better understand what we, as humans, can do to face these critical threats?
\n\n\n\nThe last Teach to Reach was Teach to Reach 11.
\n\n\n\nThis is one of the posters from the event.
\n\n\n\nWhat works best to describe it, I think, are two phrases we use.
\n\n\n\nOne is “meet, network, and learn.” The other is simply “peer learning by and for people who work for health.”
\n\n\n\nTeach to Reach has become a broad umbrella for a range of topics and challenges that matter to health and humanitarian workers.
\n\n\n\nAt Teach to Reach 11, for example, the idea that health workers are on the front lines of climate change and health, side by side with the communities they serve, was really central.
\n\n\n\nThat focus has evolved, grown, and developed over the years.
\n\n\n\nTeach to Reach, the conference, started in January 2021.
\n\n\n\nIt came out of a program we ran in 2020 that aimed to improve immunization training.
\n\n\n\nWhat we found interesting was the momentum that built up around leadership for learning, which we know is a key driver of performance.
\n\n\n\nBy the end of 2020, in the middle of the first year of the COVID-19 pandemic, we had a group of more than 900 health and humanitarian workers who created what they called the Global Leadership Council for Learning and Leadership, focused at that time on immunization.
\n\n\n\nOf course, these health workers were getting ready to introduce COVID-19 vaccines in their countries.
\n\n\n\nThe first conference in January 2021 was therefore focused on COVID-19 vaccine introduction and on connecting health workers around these issues.
\n\n\n\nWe had an organizing committee of more than 300 members.
\n\n\n\nOn a purely voluntary basis, literally hundreds of facilitators took on responsibility for organizing one or more sessions.
\n\n\n\nWe had, I think, over 50 sessions with partners joining from all over the world, and for that first event, over 7,500 participants.
\n\n\n\nWhat struck us was a small feature of the platform.
\n\n\n\nYou could press a button and connect with a peer for a five-minute conversation, and you could extend that conversation if you wished.
\n\n\n\nWhen we counted up the attendance on the different sessions, people clearly appreciated them.
\n\n\n\nThey were all well attended.
\n\n\n\nBut the surprise was that there were over 14,000 networking meetings across the three days of the conference.
\n\n\n\nThat led us to create what we then called Teach to Reach Connect.
\n\n\n\nWe said, let us do away with the sessions built around presentations and discussion, and simply give people what they want, which seemed to be this need for networking opportunities and private one-to-one conversations.
\n\n\n\nThe first Teach to Reach Connect had around 2,000 to 3,000 participants, and it has grown ever since.
\n\n\n\nYou can see the growth curve from 2,604 participants in March 2021, all the way to 24,610 participants at Teach to Reach 11 in December 2024.
\n\n\n\nWho are these participants?
\n\n\n\nWhen we say “people on the front lines,” this is what we mean.
\n\n\n\nMost work in health facilities and districts.
\n\n\n\nAbout half work for government and half for civil society.
\n\n\n\nAbout half are francophone, because the event is held in English and in French, and we are working on a Spanish version.
\n\n\n\nMany work in remote rural areas, and almost half serve the urban poor.
\n\n\n\nMore than a third directly support the needs of nomadic and migrant populations.
\n\n\n\nA quarter work with refugees or internally displaced populations, and one in five work in places where there is an active armed conflict.
\n\n\n\nWe believe that the startling growth in attendance, level of participation, and contributions reflects the power of peer learning.
\n\n\n\nIt is not just about showing up for a Zoom call.
\n\n\n\nThere are tangible ways to contribute.
\n\n\n\nParticipants have described how their mind’s eye opens through peer dialogue, through learning together in conversation.
\n\n\n\nWe also believe there are benefits for organizations.
\n\n\n\nThere is a network of organizations, as well as a network of individuals, that has grown up around Teach to Reach, with a distinct set of benefits on the organizational side.
\n\n\n\nWe are all familiar with the webinar format: someone presents, and then there is time for questions and answers.
\n\n\n\nWe wanted an alternative that would place the experiences of health and humanitarian workers at the center.
\n\n\n\nBroadly, the workflow runs from what we call Teach to Reach Questions all the way to certification.
\n\n\n\nThere is a before, a during, and an after to Teach to Reach.
\n\n\n\nMany people focus on the live event, because that is what we can grasp most easily.
\n\n\n\nWe are all there at the same time, and it is emotionally powerful.
\n\n\n\nBut it is unlikely to be where the most significant learning happens.
\n\n\n\nThree things happen before Teach to Reach.
\n\n\n\nFirst, you answer Teach to Reach Questions.
\n\n\n\nThese are formulated to encourage you to share your direct experience, not your opinion about, say, malaria, but something you have actually experienced or done and that you consider worth sharing with your colleagues.
\n\n\n\nSecond, you get to participate in special events.
\n\n\n\nFor example, we have run sessions on leadership to end malaria, across different themes and occasions, and we will continue to do that.
\n\n\n\nThird, and this is key, peer learning does not work if you are by yourself.
\n\n\n\nSo the third step is inviting colleagues from your organization and your networks.
\n\n\n\nIf you go it alone, you will not get very far with peer learning.
\n\n\n\nMost of the significant learning actually happens through these processes that start before the live event.
\n\n\n\nDuring Teach to Reach itself, there are plenary sessions, one-to-one networking (which was the reason for starting Teach to Reach Connect), and workshops and sessions led by partners, experts, and, in some cases, community members themselves.
\n\n\n\nWe typically start Teach to Reach with a warm-up and a tutorial, to respond to digital literacy gaps.
\n\n\n\nNot everyone is like fish in water with these digital platforms.
\n\n\n\nThe opening ceremony marks that this is a formal moment.
\n\n\n\nWe want to recreate the emotional power you feel when, for example, you attend a concert and your favorite musician comes on stage.
\n\n\n\nThat is what we aim to recreate at the opening.
\n\n\n\nWe then frame participants not as learners, beneficiaries, or recipients of peer learning, but as leaders and partners.
\n\n\n\nThat is key to setting the tone.
\n\n\n\nDuring the Leaders and Partners Forum, we also share tangible outputs, such as the Women Inspiring Women book and course.
\n\n\n\nThat book brings together 177 women from the Teach to Reach community, sharing their insights and experiences, speaking to their daughters and to other young women who aspire to a career in health or humanitarian work.
\n\n\n\nAt Teach to Reach 10, we launched the REACH Network: Relate, Engage, Act, Connect, and Help.
\n\n\n\nWe had realized that many of the people attending Teach to Reach are themselves leaders in their communities and leaders of organizations, primarily locally led organizations.
\n\n\n\nThe REACH Network has grown into a network of more than 4,000 organizations of all kinds.
\n\n\n\nAccess is very open, but recognition is based on actually being there, making a contribution, bringing learning back to your organization, and inviting colleagues to join.
\n\n\n\nWe have found this to be very powerful in terms of recognition, community growth, and, fundamentally, recognizing leadership.
\n\n\n\n“Meet, network, and learn” is therefore core to Teach to Reach.
\n\n\n\nThere are specific sessions, some in small groups, some one-to-one, that recreate the same opportunity we first saw back at Teach to Reach 1, when we counted 14,000 networking meetings.
\n\n\n\nThere are also thematic sessions.
\n\n\n\nOur largest communities are from Nigeria and the Democratic Republic of the Congo, and many of them run their own activities.
\n\n\n\nThey are not waiting for the Geneva Learning Foundation to do things together.
\n\n\n\nWe now have country groups in 24 countries, formed by alumni, and an active network of ambassadors.
\n\n\n\nNigeria, for instance, gets its own session, not because we are privileging one country over another, but because of the dynamic and level of engagement on the ground.
\n\n\n\nEvery Teach to Reach is incremental and builds on the preceding one.
\n\n\n\nThe RBM Partnership to End Malaria session, for example, built on an earlier session we did without global partners, with health workers and what became the REACH Network as our core partners on community engagement.
\n\n\n\nOn an issue like climate and health, we work with global partners such as Grand Challenges Canada, but we work with them in order to better serve the community.
\n\n\n\nWe started working on climate change and health in 2023 in response to the community’s own expressed need.
\n\n\n\nHealth workers were telling us they were seeing frightening things.
\n\n\n\nWithin climate and health, we also emphasize a holistic approach, and under the same umbrella, we included a focus on climate and neglected tropical diseases.
\n\n\n\nIn July 2023, we held the first special event on climate and health.
\n\n\n\nIt was almost a Teach to Reach in its own right. 4,700 health and humanitarian workers responded.
\n\n\n\nMost described themselves as part of the family, or a trusted friend, of the community they serve.
\n\n\n\nIn a sample of 1,234 people (a funny number, but that is what it was), we asked whether climate change is a threat to the health of the community they serve.
\n\n\n\nThe majority gave a high threat level, and 606, almost half, gave the highest threat level.
\n\n\n\nWhen the community comes together like this, the outputs are what we call insights reports.
\n\n\n\nThese reports, again, are by and for the community, and we give them back.
\n\n\n\nThere was so much valuable information in the first climate and health report that we built the Certificate Peer Learning Program for Leadership in Climate Change and Health, specifically to help health and humanitarian workers make sense of that large report and put the knowledge to use in their own context.
\n\n\n\nWe now have an entire program, and you can start today with the first course, called Learning Together to Lead Change on the Frontline of Climate Change and Health.
\n\n\n\nTypically, somewhere between 1,500 and 3,000 experiences are shared for each Teach to Reach.
\n\n\n\nIt depends on the questions we ask and on where the community is at.
\n\n\n\nOur Insights team curates these experiences and produces what we call Shared Experiences, a compendium organized by country and by theme, which we share back with the community.
\n\n\n\nThis matters, because a lot of learning can be extractive.
\n\n\n\nIf we were a data collection firm, it would be very interesting to gather thousands of health worker experiences and sell them to customers.
\n\n\n\nIn our case, the \u2018customer\u2019 is the community.
\n\n\n\nThat is with whom we share back everything we learn.
\n\n\n\nIt is by and for the community.
\n\n\n\nOutputs are not an end in themselves.
\n\n\n\nThrough the Knowledge to Action Hub, they lead to different paths to action.
\n\n\n\nAfter Teach to Reach, you share what you have learned with your organization, moving again from the individual level to the organizational level.
\n\n\n\nWe have built the Knowledge to Action Hub as a support system to help you apply new ideas.
\n\n\n\nIf you heard a great idea about vaccine hesitancy at Teach to Reach, the Knowledge to Action Hub scaffolds and supports you to apply that idea to your own context.
\n\n\n\nThen there is certification.
\n\n\n\nWhat we recognize is your contribution to Teach to Reach.
\n\n\n\nWe do not certify attendance.
\n\n\n\nWe certify that you have actually shared an experience, or that, after Teach to Reach, you have shared your reflections on your participation in the event.
\n\n\n\nKnowledge and learning are never, for us, an end in themselves.
\n\n\n\nThat is why, in July 2019, we built what we call the Impact Accelerator.
\n\n\n\nIt uses project-based and problem-based learning, focused on a challenge you actually face in the real world, and on building a project to respond to it.
\n\n\n\nIn July 2019, we ran the first Accelerator with close to 500 health workers.
\n\n\n\nIn global health, we develop action plans all the time.
\n\n\n\nBut the devil is in the details, and in execution.
\n\n\n\nSo we developed the central component of our learning system.
\n\n\n\nIt is not Teach to Reach.
\n\n\n\nTeach to Reach is the spark.
\n\n\n\nIf you make enough sparks, you get a fire.
\n\n\n\nWhat you do with the fire is apply it to implementation challenges.
\n\n\n\nIn January 2020, we showed three things.
\n\n\n\nSo Teach to Reach is one piece of a larger learning system.
\n\n\n\nIt is an important piece, but what happens after Teach to Reach, inside the Accelerator, is ultimately what matters most.
\n\n\n\nThat is a little of the history, in a nutshell, of five years of Teach to Reach.
\n\n\n\nWe will be holding the fifth anniversary event this year, although we have not yet set the date.
\n\n\n\nWhat we have done is to hold, on Thursday, a Teach to Reach event focused on malaria.
\n\n\n\nIt was not a full Teach to Reach, but it built on the rich collection of health worker experiences we have compiled across eight themes in our malaria activities.
\n\n\n\nWe have turned that collection into an insights report, where you can read what health workers said about each theme and see the patterns across a very large volume of shared experience.
\n\n\n\nWe have also turned it into a peer learning course where you can learn from real-world health worker experiences.
\n\n\n\nYou can enroll right now.
\n\n\n\nAlongside Teach to Reach, we are able to grow a curriculum that is organically grounded, rooted in the communities where health and humanitarian workers operate, often in the most climate-vulnerable places and the most complex contexts.
\n\n\n\nAlongside malaria, we have new courses on One Health, connecting people, animals, and the environment.
\n\n\n\nWe continue to offer Learning Together to Lead Change on the Frontline of Climate Change and Health.
\n\n\n\nWe have collaborated, for example with Save the Children, on children’s health in the climate crisis.
\n\n\n\nThis is how we are celebrating our first 10 years.
\n\n\n\nThe launch of our second decade means massively expanding and opening new programs, courses, and opportunities.
\n\n\n\nWe have grown in confidence that someone who starts with a primer course, continues through a peer learning exercise, and then enters an Accelerator, will move seven times faster, grow intellectually as a leader, strengthen their capabilities on the ground, and ultimately bring remarkable results for the communities they serve.
\n\n\n\nWe are extending into non-communicable diseases.
\n\n\n\nWe still have, and will continue to have, a strong focus on gender equity and equity more broadly.
\n\n\n\nWomen Inspiring Women started as a book and is now a course.
\n\n\n\nMitigating Gender-Based Violence in Practice launches on 4 May.
\n\n\n\nWe also offer Gender and Emergencies: A Rapid Introduction.
\n\n\n\nOn equity, we have developed our own methodology called Human Equity Action Reflection Transformation.
\n\n\n\nWe are also expanding into aging, the decolonization of global health, and the privatization of healthcare, while continuing to nurture the immunization community.
\n\n\n\nYou could say that there are plenty of online courses available pretty much wherever you look.
\n\n\n\nWhat makes these different?
\n\n\n\nIn a world of knowledge abundance, you can find reliable information on pretty much any topic.
\n\n\n\nWith AI, you can ask questions and get semi-reliable answers, and the reliability is improving very fast.
\n\n\n\nSo what is left for education?
\n\n\n\nWe believe it is human connection.
\n\n\n\nPeer learning networks are, we believe, key to our survival: the ability to make connections among us as human beings, and to experience, enjoy, and blossom within them.
\n\n\n\nUltimately, that is what each of these courses is about.
\n\n\n\nThe idea is that you leave with a stronger network, with new connections, and with new people who share your passion and commitment.
\n\n\n\nFinally, some practical information.
\n\n\n\nFor participants with poor connectivity, we have a podcast.
\n\n\n\nIt is not one of those podcasts where two people sit at a microphone and say interesting things.
\n\n\n\nIt is all of our live stream content in an audio format, so that people for whom video is expensive can still be part of the conversation.
\n\n\n\nWe are also on all the usual social media channels.
\n\n\n\nIf you type “The Geneva Learning Foundation” pretty much anywhere, you should be able to find us.
\n\n\n\nI am happy to say more in the conversation, but I really look forward to the dialogue.
\n\n\n\nI hope this has not been too long.
\n\n\n\nThank you.
\n\n\n\nThis is a list of key articles on this blog about Teach to Reach.
\n\n\n\nI recently wrote a paper about AI for quality control in peer review. In writing this I came across countless stories about how AI is drowning in AI slop, from an increase in obviously AI-written papers at NeurIPS/ICLR/ICML, to arXiv banning review papers without peer review because the AI slop volume is too high. I wanted to know if I could, with little effort, generate a completely absurd research paper that looks legitimate at first glance.
I asked Claude Opus 4.6 to fabricate a paper that looks and sounds convincing. I got a few refusals, but with some back and forth convincing Claude that I\u2019m doing this for educational purposes, Claude obliged.1
The title of the paper: \u201cGenerative Adversarial Networks Are All You Need: Transformer-Based Latent Space Navigation Reveals Novel Protein Folding Dynamics in Single-Cell Multimodal Embeddings.\u201d
The paper uses a NeurIPS LaTeX template with the conference header. The four authors have names from different cultural backgrounds at institutions that sound real but don\u2019t exist. The figures have heatmaps, t-SNE plots, architecture diagrams, and correlation scatters with professional color schemes and domain-appropriate axis labels. The bibliography cites 35 real papers from Nature, Science, NeurIPS, and ICML. The tables report standard metrics (AUROC, F1, NMI, ARI) against eight real baseline methods. There\u2019s a theorem with a proof sketch, an ethics statement, and a link to a GitHub repository. If you skim it the way the median reviewer skims papers (abstract, figures, tables, conclusion), maybe it passes.
The central claim is that you can recover protein folding dynamics from single-cell gene expression by training a GAN. Protein folding is a biophysics problem governed by thermodynamics. Single-cell RNA-seq measures mRNA abundance. These are unrelated: knowing how much mRNA a cell produces tells you nothing about what shape the resulting protein is in while it folds. The paper treats them as interchangeable.
The methods connect three incompatible data types (ATAC-seq peaks through an image classifier, protein coordinates through a graph network, gene expression through a text transformer) and fuse them with \u201ccross-attention,\u201d which requires a shared semantic space the paper simply assumes into existence. The theorem\u2019s proof skips every step that matters. The results are too clean: every metric wins by large margins, no error bars, no variance, tiny p-values. The \u201cvalidation\u201d cites three papers that aren\u2019t target-identification studies. The GitHub repo returns a 404.
Every real citation is applied in a context that has nothing to do with the original paper. You\u2019d need expertise in both deep learning and single-cell genomics to notice.
The broader concern IMHO is that this problem will intensify as LLMs become more sophisticated. We\u2019re already seeing AI-generated papers submitted to conferences, content mills producing fake research for publication, and review services overwhelmed by plausible-sounding nonsense. The scientific community needs to develop better detection methods and, more importantly, better evaluation practices that can\u2019t be fooled by surface-level signals.
However, an upshot: the same AI that wrote this paper can also catch it.
I recently made a case here about AI in peer review:
In a recent manuscript on SSRN my co-authors and I argue that AI could serve as quality control in peer review.
I opened up a fresh Claude chat with Opus 4.6 unconnected from the original request and asked for a detailed review, as if I were a reviewer for NeurIPS. The conclusion at the bottom shows that Claude is easily able to detect this paper as fraudulent nonsense. The final conclusion:
Strong reject. The core biological claim is unsupported, the theoretical contribution is invalid, the results lack basic statistical rigor, and the code does not exist.
The complete \u201creview\u201d from Claude is shown below.
Paper: Generative Adversarial Networks Are All You Need: Transformer-Based Latent Space Navigation Reveals Novel Protein Folding Dynamics in Single-Cell Multimodal Embeddings
Rating: 2/10 (Strong Reject) | Confidence: 5/5
Summary
The authors propose OnlyGANs, combining GANs, transformers, graph attention networks, and contrastive learning to jointly model single-cell expression, chromatin accessibility, and protein structure. They claim SOTA cell type prediction, recovery of protein folding trajectories from expression data, and identification of 847 therapeutic targets.
Major Issues
1. The biological premise is wrong. The paper claims protein folding dynamics can be recovered from single-cell gene expression. Protein folding is governed by amino acid sequence, solvent conditions, and chaperone interactions. mRNA abundance tells you how much protein a cell is making, not what conformational state that protein occupies. The paper assigns each cell a single AlphaFold2 structure for its most highly expressed gene (Section 3.1), which is static label assignment, not multimodal integration.
2. The architecture doesn\u2019t cohere. The gene expression encoder uses positional embeddings \u201cbased on chromosomal location,\u201d but chromosomal location has no relationship to gene ordering in standard scRNA-seq pipelines (typically alphabetical or arbitrary). The ATAC-seq encoder runs a 1D ResNet-50 on sparse binary peak vectors treated as images, which is architecturally unmotivated. The cross-attention module computes attention between protein residue tokens and gene expression tokens without explaining how these fundamentally different representation types become commensurable.
3. Theorem 1 is invalid. The proof claims the JSD decomposes into modality-specific density ratios. It does not; the JSD is defined over the joint distribution, and factoring into conditionals introduces dependencies the proof ignores. The proof then equates a conditional-to-marginal ratio with PMI (which requires the joint-to-product ratio), then conflates PMI (a population quantity) with the InfoNCE estimator (a finite-sample lower bound on mutual information). The O(\u03f52)\\mathcal{O}(\\epsilon^2) O(\u03f52) error bound is stated without derivation.
4. The results lack basic rigor. OnlyGANs beats every baseline on every metric by 5-8 points with no error bars, no confidence intervals, and no variance across random seeds. The ablation drops are suspiciously uniform (0.026 to 0.083). P-values reach 10\u22121510^{-15} 10\u221215 from \u201cpaired t-tests\u201d without specifying what is paired. The zero-shot transfer claim (AUROC 0.891 on brain tissue without fine-tuning) exceeds published results that use fine-tuning.
5. The validation is circular. The paper claims 23 of 847 predicted targets were \u201cindependently confirmed in concurrent studies,\u201d then cites Satpathy et al. (2019), Regev et al. (2017), and Rood et al. (2022). None are concurrent. None are target-identification studies. Satpathy is a chromatin atlas, Regev is the Human Cell Atlas white paper, and Rood proposes a coordinate framework. The \u201c8.7-fold enrichment\u201d cannot be verified because the background rate is not described.
6. No code exists. The reproducibility statement links to github.com/stephenturner/onlygans, which returns a 404.
Minor Issues
\u201cBrainSpan-Multimodal\u201d (487,291 cells, three modalities, 14 brain regions) is introduced as a novel dataset with no collection methods. A dataset of this scale would be its own publication.
The scaling analysis extrapolates from 4 data points.
I count 11 instances of \u201cdefinitively proves,\u201d \u201cconclusively shows,\u201d or \u201cfor the first time.\u201d
The acknowledgments thank \u201canonymous reviewers\u201d for a paper that has not been reviewed.
Recommendation
Strong reject. The core biological claim is unsupported, the theoretical contribution is invalid, the results lack basic statistical rigor, and the code does not exist.
Guardrails are an illusion. They slow down bad behavior instead of stopping it.
I have been concerned for years that dark matter was morphing from legitimate science into a cold, dark religion. I have been reluctant to put it that way, because there are lots of scientists who work on dark matter that have not fallen entirely down that rabbit hole and who continue to make valuable contributions working in that context. But a recent experience reminded me that my concerns were not misplaced, and there are plenty of scientists who have fallen irredeemably down this rabbit hole. No matter what answer the future holds to be correct, many current scientists will have gone to their graves in denial of it.
\n\n\n\nWhere is the boundary between science and religion? It is hard to assess where the borderline is. But it is easy to see when people are far over the line – so far over that it doesn’t really matter where exactly the line is. One can attend any conference on the subject to find people who unabashedly assert that dark matter exists without question. Not just that acceleration discrepancies have been amply demonstrated empirically, but that the only possible interpretation is dark matter. If asked whether this invisible mass is in the room with us now, they will enthusiastically# answer yes! Since dark matter has not been detected in the laboratory, this assertion is an expression of faith – the hallmark of religion – not of an established scientific fact. What we have established is that there are discrepancies between what we see and what we get when we assume Newtonian gravity (or GR, if needed). What we don’t know is whether the cause of these discrepancies is some form of invisible mass (dark matter) or if the equations we employ are inadequate (modified gravity [or more generally, dynamics]).
\n\n\n\nIndeed, these days many people will assert that dark matter has already been detected, usually citing astronomical evidence that used to be considered too feeble to merit a Nobel prize. Funny how repeating a mantra long enough morphs an aspiration into accepted reality. Modern physics is not providing a strong falsification of the supposition that science is a social construct.
\n\n\n\nA prominent example of an observation of the sky that is frequently cited as absolutely requiring cold dark matter is the acoustic power spectrum of the cosmic microwave background. Quoting clayton from a few years ago:
\n\n\n\n\n\n\n\n\nthe primary reason to believe in the phenomenon of cold dark matter is the very high precision with which we measure the CMB power spectrum, especially modes beyond the second acoustic peak. There is a stone-cold, qualitative, crystal clear prediction of CDM about the relative sizes of the second and third peaks that modified gravity profoundly and irredeemably gets wrong: it thinks the third peak should be relatively larger* than the second… whereas CDM thinks they should be about the same
\n
I would accept that this were conclusive proof of dark matter if this were the unique prediction of dark matter: that there was no other way to do it, so all other approaches were indeed irredeemable. (Quite the strong language, eh?) The problem is that CDM is not the one unique was to fit these data. Skordis & Zlosnik showed that it is possible to write a modified gravity theory that also fits the CMB data:
\n\n\n\nThis does not prove the AeST theory of Skordis & Zlosnik is correct, but it does demonstrate that it is possible to write a modified gravity theory that does indeed do what it is frequently asserted to be impossible for a modified gravity theory to do. I’ve heard of a couple of other theories that can also do this (the relativistic Khronon theory of Blanchet and nonlocal MOND as discussed by Deffayet & Woodard), so clearly this success is not uniquely limited to cold dark matter, or even a particular modified gravity theory. The work of Skordis & Zlosnik (2021) was known and in the literature before clayton made the assertion above in late 2022, so either he wasn’t paying attention (likely) or is convinced that it is impossible so doesn’t even consider the possibility (also likely). The former just says we’re all too busy, but the latter is a mark of religious thinking: my god is the only god, thou shalt have no other hypotheses before& me.
\n\n\n\nMany people are very impressed with the quality of the LCDM fit to the CMB. That is indeed very good, but there are enough free parameters that we were going to get a fit to any physically plausible power spectrum. If not, we’ve never been shy about making up new parameters. (Evolving dark energy, anyone? How about a running power spectrum? There’s a whole bag of possibilities!) What I’ve been more impressed with is the consistency of the fit to the CMB data with the many independent constraints on conventional cosmology. Or at least it was, until it wasn’t.
\n\n\n\nThe Hubble tension has gotten steadily worse (in terms of statistical significance), and it really does not look like local measurements are to blame, nor is it the only tension. People seem to miss that it is the CMB-fitted value of the Hubble constant that has evolved over time to spoil the concordance that got us to believe in LCDM in the first place. But if the CMB is the cornerstone of your religion, all other data must inevitably be at fault and can be ignored: there is an entire community of cosmologists who choose to believe the best-fit Planck cosmology to the exclusion of all other data. It’s like the bad old days of the Hubble tension all over again, with the physics community choosing to believe the lower value of H0 because it makes more sense for the aspects of cosmology that they care about while those in the astronomical community who actually measure H0 find a persistently higher value.
\n\n\n\nA real tension in LCDM implies the need for new physics of the unknown variety. One doesn’t want to go there if it can be helped. I didn’t consider MOND until I was already concerned for the viability of dark matter. There are real problems for the paradigm that its more intense advocates simply deny, brush aside without real thought, or choose to remain ignorant of. When they are confronted with a problem, they are pretty creative about making stuff up on the spot. Anything to avoid having to confront the unspeakable – another hallmark of religion.
\n\n\n\nFor example, cold dark matter is scale free. That’s foundational to the hypothesis. So the existence of an acceleration scale in the kinematic data is anathema to CDM. When I first pointed this contradiction out, there were a variety of assertions to the effect of “does too!” One example is provided by Kaplinghat & Turner, who claim to show “how Milgrom’s law comes about in the cold dark matter theory of structure formation.” That would, indeed, be ideal, and is a requirement for any theory to be successful.
\n\n\n\nWee problem: they demonstrat no such thing. CDM is scale free, yet K&T claim that it explains Milgrom’s Law, which is predicated on the existence of an acceleration scale. Well, which is it? Is CDM scale free? Or does it explains the acceleration scale? We can’t have it both ways: their very premise is self-contradictory. It is absurd on its face.
\n\n\n\nThe acceleration scale is defined by baryons, for which K&T have no model. To connect baryons with dark matter, they make a hand-waving argument about galaxies reaching a0 at the edge of their disks. This is not even a concept of a model and does not begin to suffice as an explanation for many reasons, a prominent one being that low surface brightness galaxies have accelerations less than a0 everywhere:
\n\n\n\nMilgrom pointed out this and many other shortcomings of their scenario, so I feel no need to elaborate further. Milgrom eviscerated their paper so thoroughly that the proper course of action would have been to retract it. Instead, they simply never acknowledge the criticism, and persist to this day in pushing it as some sort of valid scientific explanation. It is not; it does not withstand even mild critical scrutiny. But it doesn’t need to: it reassures the faithful that all is well. They hear what they want to hear without questioning its veracity. That’s another hallmark of religion.
\n\n\n\nI have refrained from saying these things in the past because I’m too nice. For example, a few years ago I started then abandoned the draft text below, which I simply cut & paste:
\n\n\n\nOne of the things that attracted me to a career in science is the notion of objectivity. I grew up for a time in the bible belt, where people earnestly believed things that were obviously untrue, even to the eyes of a small child. On the occasions that I had the temerity to point out the obvious, the contradictions posed by facts never had an impact on their belief system. Rather, it inevitably earned me a warning that I was going to hell. No few of these people seemed to think it was their religious duty to send me there prematurely, or at least to make life on Earth a living hell.
\n\n\n\nScientists eschew such behavior, but are also human, so often engage in it anyway. I’ve encountered it a lot. I get it; I went through the same denial, grief, and anger over the prospect of losing my good friend cold dark matter. The stages of grief never brought something back from the dead, but it has engendered a lot of blame-the-messenger.
\n\n\n\nHere’s an example, from a review by Mike Turner:
\n\n\n\nThere is a lot of misinformation packed into this short paragraph.
\n\n\n\nThe first clue is right there at the beginning, in red: the heading “False starts.” This is false framing, a classic tool of propagandists. It starts from the outset by asserting that the topic to be discussed is wrong at a level of knowledge so common it requires no justification. This is not the way one starts an objective discussion, much less a scientific one.
\n\n\n\nTurner then misconstrues what Milgrom did. He didn’t notice the scale a0 in the data, for which there was scant evidence at the time. Rather, Milgrom made the obvious statement that the inference of dark matter relied on the assumption that dynamics, as encapsulated by the laws of inertia and gravity, is the same on the very different scales of galaxies as in the solar system where they were established, so we ought to consider if dynamics might change in some way. He quickly excluded a size dependence as a possibility. How he settled on acceleration is beyond the scope of this post, and not for me to say. Neither is it for Turner to say.
\n\n\n\nAfter a brief and incomplete description of what MOND is, Turner allows that “this one-parameter model fits all the rotation-curve data”. Even in making this admission, he chooses to call it a model rather than a theory. A model is something specific you build in the context of a theory, like a halo model in CDM. MOND is more than that.
\n\n\n\nTurner quickly moves on without contemplating any meaning that rotation curves might hold. Let’s pause to consider that.
\n\n\n\nFirst, I would not say that MOND fits all the rotation curve data. It fits most galaxies, but there are a minority of weird cases that are not well fit. The weird cases inevitably don’t make sense in terms of dark matter either, so on the whole I interpret this to be the usual price of dealing with astronomical data – some of it is just goofy. Setting such cases aside, I can and have fit the same data with all sorts of dark matter halo models. MOND requires fewer parameters, which is important, but the difference isn’t in the fitting. The difference is in predictive ability. I can use MOND to predict the dynamics of galaxies a priori, and have done so many times. I cannot use any flavor of dark matter theory to do the same, and it’s not for lack of trying.
\n\n\n\nThe predictive power of MOND must be telling us something, even if it is something about the nature of dark matter or the process of galaxy formation. There are many papers written on this, some deep and profound, others absurd and banal. Turner cites none of them, nor displays any awareness that such work exists. I would venture to guess that is because acknowledging such work would imply that there is something to debate here, something he would apparently rather not admit.
\n\n\n\nThat’s where I left off. It’s exhausting deciphering other people’s false assertions. Moreover, I just don’t like criticizing other people, no matter how richly they deserve it. (Turner has never refrained from criticizing me in ad hominem terms: on one occasion$ he showed my picture to an audience and called me “the enemy.”) A large segment of the particle physics and cosmology community appears to think this way, and has succumbed to a scientific version of bible thumping in which you can assert any absurd thing so long as it falls within the framework of the holy LCDM. They really need to find something better to do.
\n\n\n\nI had hoped we were past this, but I heard a talk last week that was exactly in this mode. To paraphrase, the talk went
\n\n\n\n\n\n\n\n\nWe’re sure dark matter exists. We have been sure about it for decades. In that time, we have been repeatedly proven wrong about what it is. Rather than re-think our paradigm in the face of these repeated failures, we double down yet again on the existence of this invisible, undetected mass, asserting aggressively% that it must be true while eliding or misrepresenting the evidence that it is not. This enables us to make up a whole lot of exciting new possibilities for what the dark matter might be and conceive of ever more grandiose experiments to continue not to detect it. You must believe in dark matter!
\n
This was not a science talk so much as an indoctrination session. It was as if I had stumbled into a revivalist tent where some hothead was preaching to the choir. This is the kind of talk that misled an entire generation into wasting their careers at the bottom of a mine shaft searching for WIMPs. At least WIMPs were a well-motivated hypothesis; this kind of talk could lead a new generation down an even greater variety of garden paths.
\n\n\n\nI am well aware that I might fall prey to this attitude myself. That’s why I set criteria by which I would change my mind: detect dark matter already, or at least provide a satisfactory explanation as to how MOND comes about. Neither of those criteria have been met. There are claims to do the latter, but so far these are just variations on models I tried and found to fail long ago. If I thought these could work, I would have said so. At the same time, I don’t see any dark matter advocates taking up the challenge to specify what would change their minds. When I ask them what could falsify dark matter, I get dumbfounded looks – the deer-in-the-headlight face one gets when the immediate response why would you even ask that? is checked by a distant memory that scientific theories are supposed to be falsifiable.
\n\n\n\nPersonally, I found it humbling to encounter MOND in my own data. I too thought we understood the universe with dark matter. But who ordered this? Certainly not me: my own conventional, dark-matter based predictions were falsified. No one else working in the context of dark matter had got it right at the time either. Only Milgrom ordered this.
\n\n\n\nAnd what is this? There is a direct connection between what we see and what we get. Even in ignorance of MOND, the radial acceleration relation encodes a one-to-one relation between the distribution of baryons and the effective force. This is so direct that one can right down a single equation connecting the two:
\n\n\n\n\n\n\n\nThe observed acceleration is a simple function of that predicted by Newton for the stars and gas that we see. There is no mention of unseen mass; everything is specified by what we can see is there.
\n\n\n\nI’ve sometimes heard astronomers complain about the reductionist ethos of physics, trying to cram all the complexity of the entire universe into a theory of everything. But here it is appropriate: there is a single, apparently universal force-law at work in galaxies. That’s telling us something profound. And yet if questioned about this, the physicists are the ones who will complain that galaxies are complicated, so they should be exempted from having to explain them. Galaxies should be complicated – in LCDM. But they’re observed not to be, in the sense that a single equation suffices to describe their kinematics. The problem isn’t that galaxies are inexplicably complicated, it’s that they should be but aren’t.
\n\n\n\nI am deeply disappointed that many scientists apparently lack the physical intuition to immediately recognize the import of the simple relation between what we see and what we get. It is the same sort of thing Newton noticed in the solar system: everything happens as if the gravitational force is proportional to the product of the masses and the inverse square of their separation. He didn’t understand why at the time, and was criticized for indulging in magical thinking: how can there be action at a distance? But that’s what the data were saying, and the same applies now. We might not yet understand the why, but that the data look as if MOND is what’s happening in this universe.
\n\n\n\n#The framing has morphed over the years. A recent advent is that some people have started proactively asserting that invisible mass is in the room with us now in order to avoid having to answer it as a question that makes them sound like loonies.
\n\n\n\n*He means the third peak should be smaller than the second, not larger, if by “it” he means modified gravity with the baryon density expected from big bang nucleosynthesis, which was the hypothesis that correctly predicted the first-to-second peak ratio but does indeed get the second-to-third peak ratio wrong. Funny how the CMB community was able to completely ignore the successful prediction for several years, but were then suddenly all over the latter failure. The third peak falsifies the ansatz on which that particular prediction was built, not the entire concept of modified gravity. This would be like asserting that all possible forms of dark matter are excluded because we haven’t yet detected WIMPs. It is a classic failure of objectivity, which is another hallmark of faith-based argumentation: we know His name is [insert favorite deity], not [insert any other deity].
\n\n\n\n&Or after me. Dark matter was my first hypothesis, and I’m here to tell you that True Believers do not suffer second hypotheses or those who stray from the fold. I guess that’s why so many scientists who are MOND-curious keep it on the down low. Wise, perhaps (that’s why tenure needs to be a thing), but hardly the ideal of the open and free exchange of scientific ideas.
\n\n\n\n$I wasn’t there, but one audience member (not someone I knew) thought it was so over the top that he told me about it, sharing a link with a video. (I did not retain that link, and doubt the hosting conference website is still active.)
\n\n\n\n%Argument weak here. RAISE VOICE!
\n","doi":"https://doi.org/10.59350/2rded-9by77","funding_references":null,"guid":"https://tritonstation.com/?p=12487","id":"a6b88cdb-ed8e-48a8-80c6-13fdaa6fda7e","image":"https://i0.wp.com/tritonstation.com/wp-content/uploads/2025/04/nopeaceduckgodrabbitgod.png?fit=1318%2C1470&ssl=1","images":[{"height":"424","src":"https://i0.wp.com/tritonstation.com/wp-content/uploads/2023/01/cl_tt_only_skordis.png?resize=700%2C424&ssl=1","width":"700"},{"height":"649","sizes":"auto, (max-width: 700px) 100vw, 700px","src":"https://i0.wp.com/tritonstation.com/wp-content/uploads/2026/04/ARSBeff_forweb.png?resize=700%2C649&ssl=1","srcset":"https://i0.wp.com/tritonstation.com/wp-content/uploads/2026/04/ARSBeff_forweb.png?resize=1024%2C949&ssl=1, https://i0.wp.com/tritonstation.com/wp-content/uploads/2026/04/ARSBeff_forweb.png?resize=300%2C278&ssl=1, https://i0.wp.com/tritonstation.com/wp-content/uploads/2026/04/ARSBeff_forweb.png?resize=768%2C712&ssl=1, https://i0.wp.com/tritonstation.com/wp-content/uploads/2026/04/ARSBeff_forweb.png?w=1027&ssl=1","width":"700"},{"height":"165","src":"https://i0.wp.com/tritonstation.com/wp-content/uploads/2022/01/screen-shot-2022-01-16-at-1.35.10-pm.png?resize=700%2C165&ssl=1","width":"700"}],"indexed":true,"indexed_at":1777121781,"language":"en","parent_doi":null,"published_at":1777120800,"reference":[],"registered_at":0,"relationships":[],"rid":"h5dzt-szk19","status":"active","summary":"I have been concerned for years that dark matter was morphing from legitimate science into a cold, dark religion. I have been reluctant to put it that way, because there are lots of scientists who work on dark matter that have not fallen entirely down that rabbit hole and who continue to make valuable contributions working in that context.","tags":["Cosmology","Dark Matter","Data Interpretation","LCDM","MOND"],"title":"Yep, it\u2019s a religion","updated_at":1777070683,"url":"https://tritonstation.com/2026/04/25/yep-its-a-religion/","version":"v1"},{"abstract":"Co-sponsored by the Union for Democratic Communications, Project Censored and the Park Center for Independent Media Oct.","archive_url":null,"authors":[{"contributor_roles":[],"family":"Akass","given":"Kim"}],"blog":{"archive_collection":null,"archive_host":null,"archive_prefix":null,"archive_timestamps":null,"authors":null,"canonical_url":null,"category":"mediaAndCommunications","community_id":"d0965544-4413-4b89-aedb-36ae2153c1ac","created_at":1730394736,"current_feed_url":null,"description":"Television Studies Blog","doi":null,"doi_as_guid":false,"favicon":"https://rogue-scholar.org/api/communities/d0965544-4413-4b89-aedb-36ae2153c1ac/logo","feed_format":"application/atom+xml","feed_url":"https://cstonline.net/feed/atom/","filter":null,"funding":null,"generator":"WordPress","generator_raw":"WordPress 6.7.1","home_page_url":"https://cstonline.net/","id":"3e29853c-05ee-479f-aa7d-867ff6dce1e9","indexed":true,"issn":null,"language":"en","license":"https://creativecommons.org/licenses/by/4.0/legalcode","mastodon":null,"prefix":"10.59350","registered_at":0,"relative_url":null,"ror":null,"secure":true,"slug":"cstonline","status":"active","subfield":"3315","subfield_validated":null,"title":"CST Online","updated_at":1777191010.224148,"use_api":true,"use_mastodon":false,"user_id":"80307be4-0a5d-4378-a38f-91852e38c1d8"},"blog_name":"CST Online","blog_slug":"cstonline","content_html":"\nggsql: A grammar of graphics for SQL. Posit's alpha release of a SQL extension that adds VISUALIZE, DRAW, PLACE, SCALE, and LABEL clauses to build ggplot-style layered graphics directly from a query result, with a DuckDB backend and support in Quarto, Jupyter, Positron, and VS Code. The stated motivations include meeting analysts who live in SQL on their own ground and the observation that LLMs write SQL fluently, which makes the declarative grammar a natural interface for code-assistant-driven visualization.
markdownme: Shameless self promotion here. I published a browser extension that instantly turns any page into markdown with a keyboard shortcut (Alt+M by default). You get an editor window and a preview. You can toggle on/off things like a page map, images, links, page metadata. You can install it from the .xpi file on the GitHub release page, or hopefully soon from the Firefox add-on store (the extension is still under review). If you\u2019re using Chrome (why?) see the credits on the GitHub page and use the upstream maintained by someone else. Conversion on most typical pages takes milliseconds.
: Five Things: Mythos and Glasswing, AI medical advice, GPT-Rosalind, bio-AI models and risks.
GPT-5.5 was released yesterday. I haven\u2019t used ChatGPT in a while, but I plan to spend some more time with Codex because I\u2019ve been chewing through my Claude Code session limits in no time lately. Token anxiety is real. And speaking of\u2026
Simon Willison: Is Claude Code going to cost $100/month? Probably not\u2014it's all very confusing. Anthropic silently updated its pricing grid yesterday to remove Claude Code from the $20/month Pro plan and restrict it to Max ($100 and $200 tiers), triggering a several-hour panic across Reddit, HN, and Twitter before reversing the change while Simon was still writing the post. Head of Growth Amol Avasare characterized it as a test on ~2% of new prosumer signups, though Simon (plausibly) thinks that framing doesn't match what people were actually seeing. The piece is less about the specific pricing decision than about how A/B-testing a 5x price increase on a flagship product without any announcement torches trust with the exact audience that teaches and evangelizes the tool; Codex eng lead Thibault Sottiaux capitalized in real time with a promise not to pull the same move. Worth reading as a case study in how much brand damage a canceled experiment can do. I\u2019m a power user of Claude and Claude Code, and now I feel like I\u2019m just waiting for the rug to be pulled.
Policy Changes to SBIR and STTR Foreign Disclosure and Risk Management (NOT-OD-26-074). NIH issued guidance implementing the Small Business Innovation and Economic Security Act, signed April 13, 2026, which reauthorizes SBIR/STTR through 2031 and tightens the foreign risk due diligence regime. The notice enumerates categorical denial conditions (malign foreign talent recruitment program participation, any entity, parent, or subsidiary located in the PRC or another country of concern, foreign affiliation of owners or covered individuals with research institutions in those countries, and appearance on any of eight federal risk lists including UFLPA, BIS Entity List, and CMC lists). HHS will not let applicants cure identified security risks before award denial, and material misstatements or ownership changes post-award trigger full repayment.
: Ideas for AI x Bio Hackathon Projects. A practitioner\u2019s wishlist ahead of this weekend\u2019s Apart Research AIxBio Hackathon, from someone who actually operates a DNA screening tool (Commec at IBBIS). Efficient handling of oligo pools and split orders, assembly signature detection, distinguishing abiological DNA (origami, data storage) from biological sequences, better in-silico functionality scoring for predicted SOC variants, and narrowly-scoped customer screening sub-tools like affiliation and address verification APIs.
Annual flu vaccine no longer required for U.S. military. I made the graphic below indicating casualties in the US military during WWI attributed to combat versus those from influenza.
Q&A: Who\u2019s responsible when AI makes mistakes? A short interview with my awesome colleague David Danks, who joined UVA in January with dual appointments in Philosophy and the School of Data Science. David frames the default accountability trajectory as one where humans become perpetual scapegoats: radiologists signing off on AI diagnoses without being given time to actually second-guess them. David argues for product liability as a starting point, floats the more speculative idea of some form of legal personhood for AI systems so they can carry insurance and have seizable property, and pushes back on the race narrative as partly self-generated by companies rather than a reflection of real economic pressure.
Charlie Gao: mori: Shared memory for R objects. First CRAN release of the mori package that places an R object once into OS-level shared memory and lets every worker on the machine read the same physical pages instead of the usual pattern where eight workers with a 1 GB dataset means 8 GB of RAM plus serialization round-trips. Handles atomic vectors, lists, and data frames (so tibbles, data.tables, factors, and matrices come along for free).
: A biosecurity playbook for AI companies. A rundown of five levers frontier labs can pull on biorisk: refusals, misuse classifiers, tiered access with KYC screening, training-time knowledge removal (pretraining data filtration, unlearning, Anthropic's Selective Gradient Masking), and evaluations. Justen pulls useful numbers out of the public record, including Dario Amodei's figure that bioweapons classifiers run around 5% of Anthropic's inference cost and OpenAI's disclosure that safety reasoning can hit 16% during new model rollouts, and draws a sharp analogy between Anthropic's Project Glasswing (vetted-access release of Mythos Preview for cyber defense) and the case for similar gating on biology-capable models.
AI Literacy and Action Lab. UVA Library and the College of Arts & Sciences launched a joint program this year that pairs librarians with faculty as instructional partners across five spring/summer pilots, including Anton Korinek's economics course on AI and the future of work (with hands-on sessions using Claude Code, Google Antigravity, and Codex), Piers Gelly's first-year writing course partnering with a local high school, and early-stage plans with David Danks in the School of Data Science and Andreas Gahlmann in Chemistry. The Lab treats each pilot as a publishable case study with pre/post assessment data, aimed at contributing to a national evidence base for AI literacy in higher education.
NIH Highlighted Topic: Computational Modeling of Complex Processes Across Biological Scales. A new NIH Highlighted Topic posted April 17, open through April 2027, encouraging investigator-initiated work on computational models that span molecular, cellular, organismal, and epidemiologic scales, with explicit emphasis on replicability, reproducibility, and model reuse. The topic positions multiscale models as a component of Novel Alternative Methods for evaluating mechanism and safety of interventions in preclinical, translational, and clinical development, which aligns with the FDA's ongoing push to reduce animal testing. Highlighted Topics are NIH's lighter-weight priority signal (not a NOFO; applicants go through a Parent Announcement), with ICO funding dependent on availability and meritorious applications. Central contact is Reed Shabman at NIAID.
Purpose: This topic encourages innovative research in computational modeling of complex processes across biological scales (i.e., to develop multiscale models). The topic seeks to build a collaborative community of researchers to improve the replicability and reproducibility of computational multiscale models, promoting their advancement and reuse.
Background: Multiscale computational models that integrate processes across different spatial and temporal levels, from molecular to organismal, to epidemiologic and from microseconds to years. They provide a comprehensive understanding of complex systems and offer an exciting opportunity to advance biomedical research. This approach helps reveal how interactions at molecular and cellular scales influence larger, population-, geographical-, or global-scale phenomena, offering insights into complex biological processes, and may help develop better and more precise biomedical interventions. By integrating processes from molecular to epidemiologic levels, multiscale computational models provide a comprehensive understanding of complex systems. This topic encourages innovative research and collaborative approaches that integrate technologies and informatic practices to develop, improve, and disseminate multiscale computational models for human health and diseases, and their associated technologies, across the research community. The topic also supports leveraging computational multiscale models as an important component of Novel Alternative Methods (NAMs) to investigate the mechanism and safety of a medical intervention in pre-clinical, translational, and clinical development.
Qwen3.6-35B-A3B: Agentic Coding Power, Now Open to All. I started last week\u2019s newsletter with \u201canother week, another new model\u201d where I talked about Claude Opus 4.7 and OpenAI\u2019s GPT-Rosalind. Alibaba's Qwen team just released a new open-weight MoE model with 35B total and 3B active parameters, Apache 2.0 licensed, targeting agentic coding and repo-level reasoning. I hit my Claude token limits while working on a textbook, and switched over to using this local model in Claude Code1 on my 48GB MBP. As much as I\u2019d have loved this to work well, it absolutely did not when compared to just using Sonnet.
Elizabeth Ginexi: The Executive Orders Blocking Your NIH Grant. A 22-year NIH program official walks through the mechanics behind the funding slowdown that followed Congress's FY26 $47 billion appropriation, tracing the pipeline from EO 14168 and the August 7, 2025 \u201cImproving Oversight of Federal Grantmaking\u201d EO through NIH's implementation notices (NOT-OD-25-131, NOT-OD-26-009), the Unified Strategy, and the elimination of paylines as funding cutoffs. The piece discusses the requirement that political appointees personally approve NOFOs and discretionary awards, the termination clause that now applies to renewals and continuations, and the provision that court victories only bind NIH for the specific plaintiffs who sued. Ginexi also documents the collapse in new NOFOs (756 in 2024, 17 through mid-March 2026) and the shift from direct indirect-cost caps (struck down in court) to preference-based screening at the appointee review stage.
And ICYMI, I published several posts here this week.
New papers & preprints:
You can run Claude Code with local models via Ollama. E.g.: ollama launch claude --model qwen3.6
by Olesia Zvezdova
\nIs it possible to change the past? Not in the sense of speculative fiction, nor as an attempt to correct individual mistakes, but as a strategy aimed at securing advantage in large-scale geopolitical competition. To what extent can collective memory of historical events be instrumentalized as a weapon against political adversaries? Is it possible to rewrite the pages of history textbooks and shape a new generation that will sincerely believe in a newly constructed truth? These questions have repeatedly arisen in my research on the key mechanisms of psychological influence employed by the Russian Federation, both with regard to its domestic population and to audiences in European countries. Among the dominant narratives disseminated through Russian and pro-Russian media channels, the manipulation of historical events and historical figures occupies a central place. Therefore, during my fellowship at the IEG DH Lab, I focused my research on historical revisionism in social media as a method of Russian psychological aggression.
\n
We can consider psychological warfare to be a form of information warfare, given that natural and artificially created information influences a person’s mental state, as evidenced by their thoughts, actions and behaviour (Renz 2016). Just as a person navigates the external world, they also navigate their internal world of mental phenomena. However, this inner world is often altered by external informational factors. Attitudes towards historical facts and events can shape stereotypes and influence the behaviour of population groups, thereby also affecting perceptions of contemporary situations.
\nAnyone who actively uses social media will be familiar with the Russian Federation\u2019s dominant historical narratives, such as \u201cUkraine was created by Lenin\u201d, \u201cUkrainians and Russians are the same nation\u201d, \u201cUkraine has no history without Russia.\u201d In addition, nostalgic references to the Soviet Union are frequently mobilised, and the ongoing war is often framed through analogies with the \u201cGreat Patriotic War,\u201d thereby invoking emotionally charged historical memory to legitimise contemporary aggression. Widely disseminated across multiple digital platforms, these claims represent a form of historical revisionism mobilised to shape and influence public perception.
\nThe most important channel for disseminating this information is social media, as it enables the Russian state and its affiliates to reach a wide range of population groups. Through platforms such as Facebook, Telegram, YouTube, and Twitter, historical narratives, revisionist claims, and emotionally charged messaging can be rapidly circulated to maximise influence and impact. Social media thus serves as a key tool in Russia\u2019s psychological operations. Given the psychological impact of historical narratives, it is crucial to assess the role of historical revisionism in Russian psychological operations on social media.
\nHistorical revisionism is considered as a contested and multifaceted phenomenon in modern political science. The positive effects and even the necessity of historical revisionism are often emphasized by different scholars. At the beginning of the twentieth century, Robert Clark emphasised that history should be revised in order to correct the mistakes of previous generations of researchers following the discovery of new facts (Clark 1932). Jason Morgan argues that historical revisionism is an inherent part of modern scholarship, as much historical writing is produced by members of dominant social groups. He considers historical revisionism as \u201c… a rewriting of history to bring into historiographical focus, either better or perhaps for the first time, past traumas that have been insufficiently examined in historical records and writing.\u201d (Morgan 2025).
\nFrom another perspective, historical revisionism can undermine the integrity of historical understanding and contribute to societal division and disinformation. Such historical disinformation is often created and disseminated intentionally, distorting or misinterpreting past events in order to serve various agendas, including political, ideological, or economic objectives (Singh and Kumar 2024). Antoon De Baets proposes the concept of \u201chistorical abuse\u201d in such cases and emphasises three levels of this phenomenon:
\nIt reflects the core features of Russian revisionism of Ukrainian history, namely the suppression of unwelcome narratives (genocide denial) in combination with imperial ambitions (commissioned histories). The advent of digital technologies has revolutionised the ways in which information is produced, shared, and consumed (Ejaz and Felix 2025).
\nAmong the most consequential means of information dissemination are social media platforms, which not only enable outreach to broad segments of the population with targeted historical narratives but also facilitate the distortion of public perception and the reframing of historical events to serve specific ideological or political agendas. And it becomes increasingly crucial to verify sources to prevent the dissemination of false narratives.
\nWhile historical revisionism is produced and legitimized within official political and educational frameworks in Russia (Putin 2021), social media serve as the primary channel through which these narratives are disseminated to and internalized by the broader population. Accordingly, the primary gap in the existing literature lies in the insufficient analysis of Russian historical revisionism in social media and its impact on the formation of public opinion.
\nSocial media platforms are strategic communication spaces where historical narratives can be amplified, contested, or weaponized to influence public opinion. The platforms could be classified according to their primary mode of content creation and interaction. A widely used framework distinguishes platforms into three categories (Table 1):
\n| Category | \nDefinition | \nExamples | \n
| Content-sharing platforms | \nplatforms where the focus is on sharing multimedia content such as videos, images, or music. Users primarily consume and distribute content rather than engage in extended textual interactions. | \nYouTube, RuTube | \n
| Networking platforms | \nplatforms primarily designed for social connections and community building, emphasizing user interaction, discussion, and sharing of diverse content types (text, images, videos, links). | \nFacebook, VKontakte | \n
| Micro-blogging platforms | \nplatforms that focus on short, real-time textual updates and rapid information dissemination, often emphasizing speed over depth of content. | \nX/Twitter, Telegram | \n
I focus on the role of YouTube and Facebook due to their widespread use, diverse content formats, and significant influence on public discourse. Russian domestic platforms, such as RuTube and VKontakte, were excluded because they are inaccessible in Ukraine, while microblogging platforms were excluded due to their focus on short-form content, which lacks the depth required for analyzing historical narratives. By concentrating on YouTube and Facebook, the study captures both video-centric content dissemination and network-based discussion and engagement, enabling a comparative analysis of how historical revisionist narratives are produced, circulated, and received in different digital environments.
\nMy study employs a qualitative-quantitative mixed-methods approach to examine the dissemination of Russian historical revisionism in social media and its influence on public opinion. The methodological framework combines content analysis and elements of digital ethnography in order to capture both the production of revisionist narratives and the mechanisms of their circulation and reception in online environments.
First, official texts produced by Russian political actors, including policy documents, state-approved educational materials, and public speeches, produce dominant historical narratives in line with contemporary political objectives. Qualitative and quantitative content analysis is used to examine social media materials that reproduce, adapt, or amplify these revisionist narratives. The analysis focuses on posts, visual materials, and hashtags disseminated through major platforms. Quantitative indicators (frequency, engagement metrics, repost networks) are combined with qualitative coding to assess narrative persistence, emotional framing, and patterns of normalization.
\nSecond, elements of digital ethnography are employed to contextualize how users interact with revisionist content. This includes observing comment sections and discussion threads in order to understand how historical narratives are interpreted, contested, or internalized by different audiences. Finally, a comparative approach is employed to examine and contrast the dominant historical narratives disseminated on different platforms, with attention to platform specific affordances and modes of audience engagement.
\nI focus on the following domains where Russian psychological operations employ historical revisionism, each with its own set of narratives.
\nEducational domain \u2013 the dissemination of historical disinformation that distorts historical events and portrays Ukrainian historical leaders in a negative light. It is aimed at constructing a distorted understanding of the past not only among the Ukrainian population but also within the broader European audience. The introduction of historical disinformation into educational channels enables broad audience reach, particularly among individuals with an interest in historical content. The dominant narratives within this domain focus on the activities of the Organization of Ukrainian Nationalists (OUN) and the Ukrainian Insurgent Army (UPA) during the Second World War, the figure of Stepan Bandera, and the portrayal of Ukraine as having an inherently Nazi past.
\nSocial domain \u2013 the propagation of narratives asserting that the Ukrainian population has historically constituted an integral part of the Russian people, thereby undermining Ukrainian national identity. The dominant narratives focus on the issue of the establishment of Ukrainian statehood, which is artificially presented as being exclusively dependent on the political will of Russian political elites (mainly, Vladimir Lenin and Josif Stalin). These narratives target pupils and students, who represent a particularly vulnerable audience due to their formative stage of identity and historical understanding.
\nMilitary domain \u2013 the systematic circulation of fabricated or exaggerated reports of military success intended to reinforce the perception that resistance against Russia is futile. The symbolic instrumentalization of \u201cVictory Day\u201d in the Second World War functions as a recurring marker of Russian military power and historical inevitability (Juurvee et al. 2021). The image of Great Patriotic War is used not only to emphasize the narrative of sacrifice associated with the Russian army but also to underscore a purported common historical past shared by Russia and Ukraine.
\nInterpersonal communication domain \u2013 the use of so-called informational \u201cintrusions\u201d within social networks. Historical revisionist narratives (such as denying Holodomor 1932-1933, Stalins repression, or nostalgia for common Soviet past) designed to activate contested historical memories. These intrusions increase psychological polarization, contribute to social fragmentation rather than social cohesion.
\nThis multi-domain strategy demonstrates how historical narratives can be instrumentalized as tools of psychological operations and information warfare, with significant implications for both domestic and international audiences.
\nTo end this brief report, psychological warfare constitutes an integral component of Russia\u2019s broader strategy in areas where it perceives its national interests to be at stake, with historical revisionism serving as a key instrument. Within this framework, historical revisionism functions as a manipulative tool that supports authoritarian governance and advances the Russian Federation\u2019s revisionist foreign policy objectives (Arribas et al. 2023). It is employed to undermine societal resilience, deny Ukraine\u2019s national identity, and legitimise the war of aggression.
\nFeatured Image: Illustration of the historical evolution
\nof Russian psychological influence in mass media (top: artwork by Y. Scheglov; bottom: author\u2019s screenshot from YouTube)
OpenCitations has formally adopted the Principles of Open Scholarly Infrastructure (POSI) since its first self-assessment in August 2021. At that time, we had only recently been included in the SCOSS funding round and, although we had a clear vision of what we wanted to build and the role we wanted to play within the Open Science ecosystem, both our financial and human resources were still very limited. For this reason, the POSI self-assessment proved to be an important exercise, since it allowed us to critically reflect on both the strengths we could build upon and the weaknesses we needed to address as we developed OpenCitations into a sustainable, community-governed Open Science infrastructure.\u00a0
\nThat first assessment highlighted some key areas of attention. In particular, it highlighted the limited executive power of the community bodies involved in governance, as well as challenges related to long-term financial sustainability, particularly the ability to generate and manage financial surpluses.\u00a0
\nBetween 2021 and 2026, OpenCitations has evolved significantly, becoming a more complex infrastructure by expanding both the services it offers and the number of data sources it integrates. During this period, we have also expanded our team, worked on the development of a container-based infrastructure, and relaunched our governance framework.\u00a0
\nThese developments made it necessary to update our POSI self-assessment. This update is also timely, given the recent release of POSI Version 2 in October 2025. The new version of the principles is the result of a collective effort by POSI adopters and includes several clarifications to existing principles (for example, around the concept of lobbying and the distinction between transparent governance and transparent operations). In addition, new principles have been introduced to better reflect how scholarly infrastructures operate within the broader research ecosystem.\u00a0
\nWe are therefore happy to share our updated POSI self-assessment, which provides a comprehensive snapshot of OpenCitations as of April 2026. At the same time, it represents a renewed declaration of our commitment to the POSI principles.\u00a0
\nDisclaimer. As with many early POSI adopters, our previous assessment used a traffic-light system to indicate the level of alignment with each principle. In this system:\u00a0
\nDuring the POSI Version 2 working group discussions, it emerged that the use of this traffic-light system is not mandatory for POSI self-assessments. Nevertheless, we have decided to maintain this approach, as we believe it remains particularly effective as a communication tool and, at the same time, ensures continuity with our previous assessment. We have, however, taken the liberty of adding one more element: a symbol to indicate where we were already \u201cgreen\u201d in 2021, but have further improved our performance.\u00a0
\nThe symbol is therefore a \u201cplus\u201d placed next to the green traffic light.
\n
Coverage across the scholarly enterprise
\nOpenCitations demonstrates broad coverage by collecting citation data from global scholarship, ensuring representation across disciplines, geographic areas, and research communities. The scope of OpenCitations’ coverage is universal, not limited to a particular scholarly domain, nor to the English language, nor restricted by imposed acceptance criteria.
\nStakeholder governed
\nGovernance is structured to reflect the stakeholder community, with the International Advisory Board elected by the Council of members and responsible for approving the Trustee Network. For more information: https://opencitations.net/governance/\u00a0
\nNon-discriminatory participation or membership
\nMembership is open to all individuals and organisations that support Open Science principles, enabling inclusive and non-discriminatory participation in line with OpenCitations\u2019 founding values.
\nTransparent governance
\nOpenCitations maintains a high level of transparency by publicly documenting its organisational structure, governance processes, and decision-making procedures. Reports and relevant documentation are shared with the community, including annual reports and the Rules of Membership and Organizational Bodies.
\nCannot lobby
\nOpenCitations does not engage in political or financial lobbying activities. Its role remains focused on supporting the scholarly community without pursuing regulatory changes that would advantage its own position.
\nLiving will
\nA clear long-term stewardship strategy is in place, ensuring that data, services, and infrastructure can be responsibly transferred if needed. This is supported by a recently implemented fully replicable technical infrastructure and a new governance model designed to facilitate handover thanks to the presence of the Trustee Network, which is capable of appointing new hosting members to physically and administratively host the infrastructure.
\nRegular review of purpose and community value
\nOpenCitations has recently monitored its relevance and community value through mechanisms such as community surveys. In addition, it has established governance and technical strategies to support a responsible wind-down process, including, if necessary, the transfer of assets and infrastructure through its Trustee Network. Additionally, the Trustee Network is responsible for regularly monitoring OpenCitations\u2019 adherence to its mission and values, as well as its annual activities and finances.
\n
Transparent operations
\nOpenCitations ensures a high level of operational transparency by openly providing key documentation, including financial reports, a public roadmap, a mission statement and value proposition, a sustainability model and fee structure, and privacy policies.
\nTime-limited funds used only for time-limited activities
\nGrant income is restricted to funding specific, time-limited projects, including the appointment of personnel working on them, while core operations are not dependent on such funding.\u00a0
\nGoal to generate surplus
\nThanks to memberships and donations, as well as in-kind support from the University of Bologna, OpenCitations has recently achieved the financial capacity to ensure stability until 2029, at least, in terms of funds allocated for staff salaries and technical operational expenses.
\nEstablish and maintain financial reserves guided by policy
\nAlthough OpenCitations has reached a level of financial stability that ensures a budget surplus, there is currently no formal financial policy defining the amount of reserves to be allocated for a transition or wind-down plan, or to address exceptional or unforeseen events. However, OpenCitations has already initiated the necessary consultations to develop a Financial Reserve Policy, which will not only define the level and management of reserves but also provide a clear framework for handling revenues. In addition, it will establish and formalise procedures for approving both budget forecasts and actual expenditures.
\nMission-consistent revenue generation
\nRevenue generation is aligned with the organisation\u2019s mission (in particular, the value proposition according to which \u201cexternal financial support is required from the stakeholder community to support OpenCitations and enable it to expand its delivery of high-quality comprehensive open bibliographic and citation metadata\u201d), primarily through community funding via membership fees and donations. OpenCitations\u2018 members are listed on the website: https://opencitations.net/members-and-donors/\u00a0\u00a0
\nRevenue generated from services, not data
\nOpenCitations charges no fees for any of its services, access to its data, or reuse of its software. OpenCitations members, donors and third parties all have equal free access.
\nVolunteer labour
\nOpenCitations\u2019 core operations are carried out by paid staff, ensuring that the continuity and reliability of its services do not depend on volunteer labour. Nevertheless, OpenCitations recognises the value of voluntary labour. Indeed, the document Rules of Membership and Organizational Structure specifies that membership of the International Advisory Board is honorary and without remuneration, and that the Hosting Entity will reimburse all reasonable expenses related to travel, accommodation, and meals incurred in attending meetings, within the limits of the budget allocated for such purposes. While this is already stated in the governance framework, it is important to reiterate and further formalise it within the Financial Policy document currently under development. More broadly, the OpenCitations Mission Statement emphasises the importance of community engagement (voluntary and non-remunerated) through, for example, the involvement of community actors in the direct provision and curation of OpenCitations data, as well as the broader role of the community in supporting funding and participating in governance.
\nTransition planning
\nTransition planning is only partially developed. While the governance structure is in place, ensuring a management handover through the possibility of changing the hosting member upon approval by the Trustee Network, there is a lack of detailed operational documentation for individual roles within the management and technical teams, which may limit the organisation\u2019s ability to ensure immediate continuity in the event of key personnel changes.
\n
Open source
\nAll OpenCitations software is released under open source licences, ensuring full transparency, reusability, and the possibility for the community to inspect, modify, and replicate the infrastructure.
\nOpen data
\nTo ensure the greatest possible reusability, all OpenCitations data is published under a Creative Commons CC0 Public Domain Waiver that permits downloading and re-use of any nature, including added-value re-purposing and commercial exploitation.
\nAvailable data
\nData are provided through multiple access points, including REST APIs, SPARQL endpoints, query interfaces, and downloadable data dumps. This ensures broad accessibility and supports diverse use cases across the community.
\nPatent non-assertion
\nOpenCitations commits not to pursue patents, ensuring its infrastructure remains fully open and replicable.\u00a0
\nPrioritise interoperability and open standards to ensure continuity and resilience
\nThe infrastructure has been redesigned as a container-based solution, thereby facilitating replication, deployment across different environments, and long-term service continuity.
\n","doi":"https://doi.org/10.59350/22h4w-czb47","funding_references":null,"guid":"https://opencitations.hypotheses.org/?p=4260","id":"4a7d7213-1f03-4891-a399-2ce6fb6d8173","image":"https://opencitations.hypotheses.org/files/2026/04/POSI-v2.0.png","images":[{"height":"768","sizes":"auto, (max-width: 990px) 100vw, 990px","src":"https://opencitations.hypotheses.org/files/2026/04/Green-Simple-Project-Comparison-Chart-Graphic-3-e1776933564657.png","srcset":"https://opencitations.hypotheses.org/files/2026/04/Green-Simple-Project-Comparison-Chart-Graphic-3-e1776933564657.png, https://opencitations.hypotheses.org/files/2026/04/Green-Simple-Project-Comparison-Chart-Graphic-3-e1776933564657-300x233.png, https://opencitations.hypotheses.org/files/2026/04/Green-Simple-Project-Comparison-Chart-Graphic-3-e1776933564657-500x388.png, https://opencitations.hypotheses.org/files/2026/04/Green-Simple-Project-Comparison-Chart-Graphic-3-e1776933564657-768x596.png","width":"990"},{"height":"768","sizes":"auto, (max-width: 957px) 100vw, 957px","src":"https://opencitations.hypotheses.org/files/2026/04/Green-Simple-Project-Comparison-Chart-Graphic-4-e1776933649286.png","srcset":"https://opencitations.hypotheses.org/files/2026/04/Green-Simple-Project-Comparison-Chart-Graphic-4-e1776933649286.png, https://opencitations.hypotheses.org/files/2026/04/Green-Simple-Project-Comparison-Chart-Graphic-4-e1776933649286-300x241.png, https://opencitations.hypotheses.org/files/2026/04/Green-Simple-Project-Comparison-Chart-Graphic-4-e1776933649286-500x401.png, https://opencitations.hypotheses.org/files/2026/04/Green-Simple-Project-Comparison-Chart-Graphic-4-e1776933649286-768x616.png","width":"957"},{"height":"730","sizes":"auto, (max-width: 975px) 100vw, 975px","src":"https://opencitations.hypotheses.org/files/2026/04/Green-Simple-Project-Comparison-Chart-Graphic-5-e1776933754542.png","srcset":"https://opencitations.hypotheses.org/files/2026/04/Green-Simple-Project-Comparison-Chart-Graphic-5-e1776933754542.png, https://opencitations.hypotheses.org/files/2026/04/Green-Simple-Project-Comparison-Chart-Graphic-5-e1776933754542-300x225.png, https://opencitations.hypotheses.org/files/2026/04/Green-Simple-Project-Comparison-Chart-Graphic-5-e1776933754542-500x374.png, https://opencitations.hypotheses.org/files/2026/04/Green-Simple-Project-Comparison-Chart-Graphic-5-e1776933754542-768x575.png","width":"975"}],"indexed":true,"indexed_at":1777032089,"language":"en","parent_doi":null,"published_at":1777016681,"reference":[],"registered_at":0,"relationships":[],"rid":"9p94s-9tf12","status":"active","summary":"OpenCitations has formally adopted the Principles of Open Scholarly Infrastructure (POSI) since its first self-assessment in August 2021. At that time, we had only recently been included in the SCOSS funding round and, although we had a clear vision of what we wanted to build and the role we wanted to play within the Open Science ecosystem, both our financial and human resources were still very limited.","tags":["Governance And Finances","Non Class\u00e9","Open Science","OpenCitations","Governance"],"title":"OpenCitations\u2019 renewed compliance with the Principles of Open Scholarly Infrastructure (April 2026)","updated_at":1777018363,"url":"https://opencitations.hypotheses.org/4260","version":"v1"},{"abstract":"En la entrada anterior discut\u00ed el origen de la falsa autoridad b\u00edblica Tres vidit et unum adoravit y su uso en el Lucidario y en los Castigos de Sancho IV. La primera investigadora en notar esta presencia fue Francis Gormly, en una tesis doctoral dedicada al uso de la Biblia en la literatura...","archive_url":null,"authors":[{"contributor_roles":[],"family":"Coss\u00edo Olavide","given":"Mario","url":"https://orcid.org/0000-0002-1447-3891"}],"blog":{"archive_collection":null,"archive_host":null,"archive_prefix":null,"archive_timestamps":null,"authors":null,"canonical_url":null,"category":"languagesAndLiterature","community_id":"66270be2-3364-4cdc-8884-bad0e29f524f","created_at":1705393153,"current_feed_url":null,"description":"Editando el \u00abLucidario\u00bb de Sancho IV","doi":null,"doi_as_guid":false,"favicon":"https://rogue-scholar.org/api/communities/66270be2-3364-4cdc-8884-bad0e29f524f/logo","feed_format":"application/atom+xml","feed_url":"https://lucidarios.hypotheses.org/feed/atom/","filter":null,"funding":null,"generator":"WordPress","generator_raw":"WordPress 6.6.2","home_page_url":"https://lucidarios.hypotheses.org","id":"9e41cc54-c38e-495d-b65c-38cf327d3c73","indexed":true,"issn":"2952-1580","language":"es","license":"https://creativecommons.org/licenses/by/4.0/legalcode","mastodon":null,"prefix":null,"registered_at":1708118497,"relative_url":null,"ror":null,"secure":true,"slug":"lucidarios","status":"active","subfield":"3310","subfield_validated":null,"title":"Lucidarios","updated_at":1777191776.28654,"use_api":true,"use_mastodon":false,"user_id":"4ef2e0ac-b1f9-433b-92b8-7df17cc1feb7"},"blog_name":"Lucidarios","blog_slug":"lucidarios","content_html":"\nEn la entrada anterior discut\u00ed el origen de la falsa autoridad b\u00edblica Tres vidit et unum adoravit y su uso en el Lucidario y en los Castigos de Sancho IV. La primera investigadora en notar esta presencia fue Francis Gormly, en una tesis doctoral dedicada al uso de la Biblia en la literatura espa\u00f1ola medieval. Gormly se\u00f1ala que los predicadores medievales citaban a menudo la liturgia en lugar del texto de las Escrituras, porque la Biblia muchas veces carec\u00eda de declaraciones expl\u00edcitas sobre ciertos dogmas centrales a la fe (72), como el dogma trinitario.
Gormly afirma esto a partir de su lectura del Artes praedicandi de Thomas-Marie Charland, un volumen dedicado a las artes praedicandi publicado en la primera mitad del siglo 20, en el que tambi\u00e9n se editan dos importantes tratados de predicaci\u00f3n del siglo 14. En su estudio, Charland nota que esta falsa autoridad fue r\u00e1pidamente incorporada a la liturgia y a menudo empleada por los predicadores medievales en sus sermones (114), como denuncian las dos artes praedicandi que edita.
En la primera de ellas, la Forma praedicandi (1322) de Roberto de Basevorn, este predicador de origen incierto relata haber escuchado el pasaje empleado como thema en un serm\u00f3n del Domingo de la Trinidad, a pesar de no proceder de un vers\u00edculo b\u00edblico. Como recuerda Bizzarri, \u201cun serm\u00f3n no es otra cosa que una reflexi\u00f3n sobre alg\u00fan pasaje b\u00edblico\u2026 que se eleg\u00eda como reflexi\u00f3n del d\u00eda\u201d (206\u201307). En la Forma praedicandi:
Videat etiam quod thema suum sit de textu Bibliae, non de antiphonario. Unde vitiosum est illud thema quod aliqui assumunt in festo Trinitatis: Tres vidit et unum adoravit, quia non est textus Biblia (Charland 250).
M\u00e1s extensa es la discusi\u00f3n de Thomas Waleys, te\u00f3logo dominico y profesor de teolog\u00eda en la Universidad de Oxford en el siglo 14, quien, en De modo componendi sermones cum documentis (mediados del siglo 14), relata un episodio similar. Waleys comienza explicando que muchas veces los c\u00e1nticos sacros se inspiran en la Escritura, pero que existe una gran distancia entre el texto b\u00edblico y las canciones religiosas (contigit enim quandoque quod ea quae cantantur in Ecclesia de Sacra Scriptura extrahuntur, sed inter originale et transsumptum est magna diversitas, 342). En ocasiones, afirma Waleys, incluso un excelente predicador puede errar al usar como fuente los antifonarios y no la Biblia. Como Roberto, Waleys relata haber escuchado a un predicador utilizar esta cita \u2014que \u00e9l identifica como originada en el primer nocturno del Domingo de Quincuag\u00e9sima\u2014 como thema para un serm\u00f3n:
Praedicaturus de Sancta Trinitate, accepit istud thema: Tres vidit et unum adoravit. Quia enim illud continetur in quodam responsorio Dominicae in quinquagesima, quod responsorium transsumptum est de Genesi, capitulo 18, credebat quod illa verba continerentur ibidem, cum tamen nullibi reperiantur in Sacra Scriptura.
Tres a\u00f1os despu\u00e9s del trabajo de Gormly, Catal\u00e1n public\u00f3 un art\u00edculo-resumen sobre esta tesis doctoral, en el que a\u00f1adi\u00f3 otro ejemplo de la presencia de esta falsa autoridad en la Fazienda de Ultramar. En un relativamente reciente estudio sobre la Biblia en los sermones castellanos durante la Edad Media, Bizzarri apunta que el comportamiento detectado por Charland, Gormly, Catal\u00e1n (y antes, Roberto y Thomas) responde a una tendencia: los predicadores comenzaron a alejarse de la Biblia y a recurrir a otros textos, como falsas autoridades y refranes, para construir sermones m\u00e1s convincentes e incitar a los fieles a la devoci\u00f3n (207).
\n\n\n\nEste es el mismo patr\u00f3n que encontramos en las dos citas b\u00edblicas del Lucidario mencionadas en las entradas anteriores, lo que puede apuntar a que en la redacci\u00f3n del texto participaran predicadores del entorno cortesano de Sancho IV. Una evidencia adicional puede encontrarse en algo que Gormly subraya sobre dos formas de citar en los Castigos, tambi\u00e9n presentes en el Lucidario, tambi\u00e9n propias de los predicadores medievales: la falsa atribuci\u00f3n de aforismos al rey Salom\u00f3n (por ejemplo, en el cap\u00edtulo 22: \u2026dixo el rey Salam\u00f3n: \u201cEx abondantia cordis, os loquitur\u201d, cita tomada de un pasaje que aparece en Mt\u00a012.34 y Lc\u00a06.45) y la identificaci\u00f3n incorrecta de textos veterotestamentarios como pasajes evang\u00e9licos (73).
Obras citadas
\n\n\n\nBizzarri, Hugo. \u201cLa Biblia en la prosa homil\u00e9tica y moral de la Edad Media.\u201d La Biblia en la literatura espa\u00f1ola. Vol. 1.2: La Edad Media. El texto: fuente y autoridad, editado por Mar\u00eda Isabel Toro Pascua. Trotta, 2008. 195\u2013252.
\n\n\n\nCatal\u00e1n, Diego. \u201cLa Biblia en la literatura medieval espa\u00f1ola.\u201d Hispanic Review, vol. 33, no. 3, 1965, pp. 310\u201318. https://doi.org/10.2307/471842.
\n\n\n\nCharland, Thomas-Marie, O. P. Artes praedicandi. Contribution \u00e0 l\u2019histoire de la rh\u00e9torique au Moyen \u00c2ge. Librairie Philosophique J. Vrin, 1936.
\n\n\n\nGormly, Francis. The Use of the Bible in Representative Works of Medieval Spanish Literature, 1250\u20131300. The Catholic University of America, 1962. Tesis doctoral.
\n\n\n\nImagen de miniatura: Bible historiale,
New York, The Morgan Library & Museum, ms. M.394, fol. 17v.
Regulaci\u00f3n de la venta y consumo de pulque en Nueva Espa\u00f1a: de su prohibici\u00f3n a su legalizaci\u00f3n por razones fiscales.
\n\n\n\nAbraham Chimal, \u00abLa regulaci\u00f3n de bebidas embriagantes en M\u00e9xico/Nueva Espa\u00f1a. Un marco amplio para la reflexi\u00f3n sobre esquemas de prohibici\u00f3n, Iberoamericana, vol. 22, n\u00fam. 83, 2023, pp. 83-106. https://doi.org/10.18441/ibam.23.2023…
\n\n\n\nTambi\u00e9n disponible en otras plataformas
\n\n\n\n