Newsletter
SimPPL's 2025 Wrapped and 2026 Plans
As we step into 2026, we are reflecting on a year of building, learning, and momentum at SimPPL.
In 2025, we reached several important milestones:
- Received ~3.9k visitors to our site growing 75% within the first six months this year!
- Grew our software and research engineering teams to double the size!
- Launched the SimPPL newsletter to 200+ subscribers growing 2.3% month on month
- Released a beta for our flagship product, Arbiter scaling it to trace digital narratives in over 20 million social media posts a week!
- Completed our GenAI health-literacy pilot with 100+ hundred families via Sakhi piloted in Jalgaon and 200+ families via Maitri piloted in Bangladesh,
- Launched a rigorous, trilingual LLM evaluation benchmark on 1000 reproductive healthcare questions focused on accuracy, linguistic quality, and safety for rural India.
We expanded and launched projects with Spreeha Foundation in Bangladesh, Deutsche Welle Akademie in Germany, Southeast Asia, and Kenya, Cohere AI based in North America, and with Jagran New Media in India. We won awards from Google, Omidyar Network, Ford Foundation, and several others jointly with our partners at Harvard University and MIT.
The year also ended on a particularly high note.
A special thanks to Sharda University School of Media, Film, and Entertainment in collaboration with Jagran New Media for featuring a talk about our work on information integrity.
Our contributions were recognized by the Collective Intelligence Project for our developing one of the first LLM evaluations specific to multilingual reproductive health conversations, assessing medical accuracy, linguistic quality, and safety. This work highlights critical risks in real world deployments as AI-mediated health guidance scales. We extend a special thanks to Faisal Lalani from CIP for his collaboration and support!
In 2026, we are excited about what's ahead: new Arbiter case studies developed with journalists exploring how AI discourse differs across African countries, how H-1B conversations reflect broader anxieties about labor and migration, and how digital governance debates are evolving as governments move to hold Big Tech accountable; along with deeper research partnerships and continued work at the intersection of AI, media, and public interest.
Keep an eye out for future newsletters as we share more from the year ahead at SimPPL.
Builds, bugs, and breakthroughs
deepdarkCTI/twitter_threat_actors.md at main · fastfire/deepdarkCTI
Collection of Cyber Threat Intelligence sources from the deep and dark web - fastfire/deepdarkCTI
Learn MoreReciprocal rank fusion | Reference
Reciprocal rank fusion (RRF) is a method for combining multiple result sets with different relevance indicators into a single result set. RRF requires...
Learn MoreGitHub - aiming-lab/SimpleMem: SimpleMem: Efficient Lifelong Memory...
SimpleMem: Efficient Lifelong Memory for LLM Agents - aiming-lab/SimpleMem
Learn MoreArticles we are obsessed with
Policy Implications of China's Open-Weight AI Ecosystem | Stanf...
🆕 Have Chinese AI models pulled ahead of their global counterparts? Our latest issue brief analyzes China’s diverse open-weight model ecosystem and examines the policy implications of their widespread global diffusion. Read the key highlights and full report here: https://lnkd.in/g5Wvx68S This brief was authored by Caroline Meinhardt, S...
Learn MoreThe Case for Designing Tech for Social Cohesion: The Limits of Cont...
Based on nearly 60 interviews with staff at tech companies, critics of big tech, civil society groups impacted by tech-amplified social media, and new tech startups, research revealed three distinct but complementary narratives or approaches to thinking about polarization and social cohesion in digital spaces. The “User-Centered” Narrative d...
Learn MoreUnpacking deceptive design
Dark patterns – deceptive designs that steer users into unintended actions – erode the trust that is essential to a healthy internet. To better understand this problem, we moved beyond rigid technical definitions to ask a simple question: How do people actually experience and perceive these designs?
Learn MorePCIO Platform Interventions Codebook
Platform Interventions Literature Review Codebook Platform Interventions: How Social Media Counters Influence Operations Original created: July 22, 2020 This version created: December 10, 2020 Author: Kamya Yadav Field Description Name/Short description If the entry is a platform interv...
Learn MoreCool research to follow
Eliciting Frontier Model Character Training
A study of personality convergence across language models
Learn MoreQalb: Largest State-of-the-Art Urdu Large Language Model for 230M S...
Despite remarkable progress in large language models, Urdu—a language spoken by over 230 million people—remains critically underrepresented in modern NLP systems. We introduce Qalb, achieving state-of-the-art performance with a weighted average score of 90.34.
Learn MoreTraining AI Co-Scientists Using Rubric Rewards
View recent discussion. Abstract: AI co-scientists are emerging as a tool to assist human researchers in achieving their research goals. A crucial feature of these AI co-scientists is the ability to generate a research plan given a set of aims and constraints. The plan may be used by researchers for brainstorming, or may even be implemented afte...
Learn MoreArtificial intelligence tools expand scientists’ impact but contr...
Nature - Artificial intelligence boosts individual scientists’ output, citations and career progression, but collectively narrows research diversity and reduces collaboration, concentrating...
Learn MoreUseful data drops
In collaboration with Nature, I investigated the impact of the Trum...
In collaboration with Nature, I investigated the impact of the Trump administration on US science one year after its return to office. More than 7,800 research grants were cancelled or frozen, affecting around 25,000 scientists and research staff and resulting in an estimated US$32 billion in lost funding. This project asks a methodological qu...
Learn MoreDatawrapper: Create charts, maps, and tables
Create interactive, responsive & beautiful data visualizations with the online tool Datawrapper — no code required.
Learn MoreUS science after a year of Trump: what has been lost and what remains
A series of graphics reveals how the Trump administration has sought historic cuts to science and the research workforce.
Learn MoreShow HN: Self-host Reddit – 2.38B posts, works offline, yours forever
Reddit's API is effectively dead for archival. Third-party apps are gone. Reddit has threatened to cut off access to the Pushshift dataset multiple times. But 3.28TB of Reddit history exists as a torrent right now, and I built a tool to turn it into something you can browse on your own hardware.The key point: This doesn't touch Reddit's servers....
Learn More