Metagov News (Special AI Issue) - Nov 2025
Project updates, upcoming events, and curated readings straight from Metagov’s AI-related research areas.
Hello Metagovernors!
Everyone’s got a lot to say about AI these days… We know that it can feel overwhelming, confusing, and perhaps even worrisome 🤯
But today, we hope to offer some glimmers of hope in the form of highly-curated, multidisciplinary research-backed AI-related project updates and readings straight from Metagov’s leadership.
The Overview:
🎉 Public AI Project Updates
🗓️ Upcoming Virtual Event: Opportunities for Collective Governance of AI (Zine-Making) - Thurs Nov 13
🧠 Thoughtful & thought-provoking AI-related essays by Metagov RDs (summarized below)
Ready to dive in? Let’s slow down—perhaps brew a cup of tea and find a comfortable seat. Now, come journey with us through this edition of Metagov News (Special AI Issue). We hope you find something special here. ☕️✨👀
🎉 Public AI Updates
Public AI is one of Metagov’s fiscally hosted projects. It is led by Research Directors Josh Tan and Nick Vincent with many other contributors and affiliates that make up the Public AI Network. In short, Public AI envisions artificial intelligence as a shared public resource—developed, owned, and governed by the public, much like electricity, libraries, or the internet. This marks a shift from privately-owned AI systems toward democratic and nonprofit ownership—treating AI as a global commons.
On September 2, 2025, Public AI Network launched The Public AI Inference Utility: the public access point for public and sovereign AI models.
The Utility uses a fully-featured, open-source frontend and a deployment layer that runs on compute from public and private partners around the world. We offer stable, direct access to models built by national (and international) public institutions. Imagine a water or electric utility, but instead of H20 or electrons, you’re getting inference on tap.
You can try the Inference Utility at publicai.co and access multiple vetted models including Apertus from Switzerland, SEA-LION v4 from Singapore, and more coming soon. See this official press release for more details about how Swiss AI’s Apertus language model is now available internationally through the Inference Utility. Huge shoutout and congratulations to the Public AI team—especially Joseph Low, research engineer/technical lead and Metagov community member! 🎉
Why is this special? Right now, the Public AI Inference Utility is the only public inference provider on HF... meaning that our inference is FREE. To be clear, a lot of donated compute goes into this, but we are working our butts off to keep that inference free (or at least extremely cheap) for developers, researchers, and businesses that want to build on it.
Be sure to subscribe to the Public AI Newsletter for all project updates; see about Contributing; and/or email hello@publicai.co if you are a lab or government with a model that you want to host on Hugging Face.
🗓️ Upcoming Event 11/13: Opportunities for Collective Governance of AI (Zine-Making)
Last month, Research Director Nathan Schneider shared a document he’d begun in the Metagov Slack titled “AI and Collective Governance: Points of Intervention” and invited community members to contribute examples of opportunities for collective governance at each layer of the AI stack including: model design, data, training, hosting, user experience, and more.1
Nathan has since been incorporating a variety of contributions to the resource and next Thursday, November 13, 2025, will be leading a (virtual) collaborative zine-making session where we can come together synchronously to discuss the rationale and purpose behind the effort and further enrich the resource. We’d especially love creative folks passionate about visual design for zines and other collective publications to join & inspire us!
🧠 Research Dispatches: Two Essays from Metagov RDs
1. “Meltdown” by Michael Zargham and Ilan Ben-Meir
Summary
“Meltdown” argues that what we need most in today’s AI landscape—where generative systems are producing text, code, images, and decisions at previously-unimaginable speeds and scales—is complementary systems to direct, constrain, validate, and co-regulate these systems: this is “regulative AI.”
Online and offline, human societies are structured around protocols and institutions such as markets, states, firms, and legal systems that help us coordinate our actions at scale. Without analogous systems that interpret and validate generative AI production, the technology risks overwhelming human capacity for discernment.
“Generative AI produces information so efficiently that it has overwhelmed humanity’s ability to manage that production effectively…. At this point, we do not need systems that are better at producing information, but rather systems that can interpret, coordinate, verify, and validate such production in service of specific aims.”
Read the full piece for more information about how a return to the foundations of cybernetics (signal processing, control theory, and systems engineering) can help us avert the present meltdown.
2. “Developmental Artificial Intelligence: When the job of AI is to make itself obsolete” by Seth Frey
Summary:
Seth’s piece explores the question: “What kind of work should we favor for AIs, and which should we work to keep to ourselves?” He makes the distinction between formative work (work designed to shape, develop, or mold a person) vs. summative work (work designed to produce a specific outcome), and argues that perhaps we’ll find more social agreement to automate the latter whereas automating the former will cause the decay of fundamental skills. In fact, the role of tools (including AI) in formative work, Seth argues, should be to make themselves obsolete.
He uses this framework to analyze specific jobs including note-taking, art-making, education, and governance.
“Governance is especially timely and relevant. I’m increasingly obsessed with governance as a thing that people use to develop themselves. So I’m nervous about AI facilitation, argumentation, and deliberation because they are developed by people who assume that all governance work is summative. I believe more and more that a surprising chunk of it is formative, and AI will make democracy worse as our good habits decay.”
In other words, in the domain of civic deliberation, the work of facilitation is formative (important for human development). How could AI be used in this field? Seth writes, “Instead of using human discourse datasets to train AI facilitators, we should use them to train AI debators that we use to train human facilitators.” Metagov actively works on research and tool-building for deliberation facilitation—a research area led by RD and Metagov ED Liz Barry—because of the concerns expressed by Seth around some worrisome AI implementations in this domain.
Seth expresses skepticism of tool providers—including AI service providers—in domains of formative work because of the business’ inherent incentives for profit and longevity. Therefore, he makes the case for public AI because “no privately held bot is trying to make itself obsolete.”
Other AI-Related Metagov Updates
The Collective Intelligence Project launched Weval: The Open Platform for AI Evaluation! Create, share, and run a massive, collaborative library of AI model evaluations. Like a Wikipedia for benchmarks, Weval empowers anyone to test what matters to them. Congrats Metagov RD Divya Siddarth & the rest of the CIP team! 🥳
This past summer, we started a Metagov Book Club with the first book being Empire of AI by Karen Hao. If you’ve read the book and are interested in reading our discussion notes or joining ongoing conversations, join us on the Metagov Slack and reach out!
Upcoming Metagov Events
Monthly Orientation
Wednesday 11/12: Join us for one of our Metagov Community Orientation Calls that we host every month. Come to learn about Metagov’s mission and approach as a laboratory for digital governance, meet fellow new members of the community, and learn about ways to get involved in research or other community activities.
Metagov Seminars: We have no seminar scheduled for this Wednesday, but next week we have 2x seminars!
Monday 11/10: “Open Facilitation Library Seminar with Stanford Deliberative Democracy Lab“ where we’ll hear from the team behind Stanford’s innovative video platform for large-scale deliberative polling.
Wednesday 11/12: “Governance Among Plants & Insects with Dr. Orit Peleg“ which will seek to understand behavior of disordered living systems by merging tools from physics, biology, engineering, and computer science
Community Calls:
This Thursday 11/6: meeting of the Metagov book club where we will be discussing the full book: The Dispossessed by Ursula Le Guin. Can’t wait to dig into this one with many of you. If you have ever read this book (even if not this month), please join us!
Next Thursday 11/13: Zine-Making Session for “AI and Collective Governance: Points of Intervention hosted by Nathan Schneider who has put together a participatory project to collect examples of collective governance of the AI stack.
NOW SEEKING: Metagov Playtesters
We are looking to create an official group of Metagov Playtesters who want to be invited to test out the wide variety of tools - from deliberation, to interpersonal conflict resolution, to educational tools that folks in our community are working on! Does this sound like you? To become an official Metagov Playtester, just sign up for any event on our Luma with “Playtest” in the title, and you will automatically be added to this go-to group for future opportunities. This means you are a trusted source of feedback for the tool-builders in the community, so please sign up if you are curious, open-minded, and willing to provide constructive feedback to these project teams.
This Friday (and every Friday in November), our Governable Spacemakers Fellowship team will be hosting Playtest sessions for their Small Hassles Court mini game. Sign up for any slot (11/7, 11/14, 11/21, 11/28) and become an official member of the Metagov Playtester community.
Past Seminars
9/10: “Understanding Early Large-Scale Collectives” (ft. Justin Jennings, Barbara Mills, & Gary Feinman). Watch on Youtube or the Internet Archive.
9/24: Metagov x Future of Science Seminar: “Scaling Digital Support Systems” with David Sisson. Watch on Youtube or Internet Archive.
10/8: “Other Networks” with Dr. Lori Emerson. Watch on Youtube or IA.
10/22: “Beyond Governance: Designing for Decentralized Futures with Indy Johar.” Watch on Youtube or IA.
10/29: “Online Governance Surfaces and Attention Economy” with Nathan Schneider, Michael Zargham, Ronen Tamari, Kelsie Nabben. Watch on Youtube or IA.
There are so many ways to get involved, meet people, collaborate, and connect in the Metagov community. If you have any questions, reach out to Metagov’s Community Lead, Val Elefante, on Slack. Thanks for being here =D
Side note: this practice of open document sharing and collective writing has become fairly common in the Metagov community and has led to some rich, community-developed resources the Notes on Governance Experience Design piece as another example.