Digest Issue 72: Introducing GreenEarth | Enhancing Trust in Agentic Commerce | The Embedded Economy | Saving democracy for the price of a swimming pool

Curated collection of media on the interactions between humans and technology and what it means for the future of the human-technology relationship

Photo by sq lim on Unsplash

Hi Distroidians!

I hope you are all doing well.

For this issue, I included additional information for some of the curated media beyond simply the Title and URL, with the use of a large language model (LLM) for the Takeaways.

Let us know your thoughts in the Comments section on this issue as we continue to iterate on the Digest.

Digest

Agency in the New Age of Communication

• Date: 2025-10-21

• Author: TBA

Introducing GreenEarth

• Date: 2025-10-23

• Author: Jonathan Stray

Description

We’re building advanced open source algorithms for social media

Summary

Key features on our roadmap include interactive customization through LLM prompts (“show only the funniest cat videos”), engagement prediction models, and selectable candidate generators. There are many ways to do this, from just collecting every post from the people you follow, to searching the entire inventory for topics that you’ve previously liked, or analyzing your social network (Twitter’s source code includes several). The set of candidate generators, prompts, scoring weights, and reranking rules is a design space that we are hoping many people will be able to create wonderful things with.

Takeaways

1. GreenEarth addresses complaints about social media feeds by offering open, customizable recommendation technology for ATProto.

2. Key features include interactive LLM prompt customization, engagement prediction models, and selectable candidate generators.

3. The system allows deep user configuration of every recommendation pipeline step (candidate gen, scoring, reranking).

4. Infrastructure will be available as both open-source and a paid API service.

5. An MVP recommender system is planned for release in Q2 2026.

6. The project is led by Renee DiResta, Glen Weyl, Jonathan Stray, Ian Baker, and Nikhil Garg, funded by a $300,000 seed grant from Project Liberty.

7. Individuals can get involved by joining their Discord server or contributing to their GitHub repository.

Historic: The Inaugural AI Agent Congress in Silicon Valley - Jeremiah Owyang

• Date: 2024-11-24

• Author: Jeremiah Owyang

Description

An industry-wide event of the leading AI Agent CEOs discussing and planning at industry-level. The event format was unique, moving from structured to unstructured: 1) A formal industry presentation at the start, followed by 2) facilitated group discussions on three topics, and finally 3) an unconference, where attendees (called delegates) could crowdsource topics and attend the sessions of their choice.Historic:…

Takeaways

Takeaways:

1. The inaugural AI Agent Congress successfully connected key players in the emerging AI agent market

2. Industry debate continues about developing one meta-agent versus a swarm of specialized agents

3. Open protocols and standards are needed to ensure interoperability across different AI agents

4. AI startups are expected to outmaneuver big tech companies in the AI agent space

5. The event’s evolving format fostered effective collaboration and knowledge sharing

6. An ongoing community has been formed to continue discussions and partnerships

7. The next AI Agent Congress is planned for the first half of 2025

Enhancing Trust in Agentic Commerce

• Date: 2025-09-18

• Author: Ginny Fahs

Description

CR joins Skyfire at Visa’s SF HQ to present two Agentic Commerce prototypes, envisioning shopping agents that guide consumers with trust

Summary

We’re seeing increasing interest in “agentic commerce”: shopping experiences through which consumers rely on AI-powered tools to not only help them research products, but also make purchase decisions and carry out transactions. We’re already thinking about how to get this functionality to market, as it will require new guardrails, consent frameworks, patterns for agent authorization and permissioning, and novel experience design to work in a production setting. Skyfire once again powers the identity and payment layer, while CR’s role is to stream our verified data to the agent in order to more effectively guide and advise the consumer at the point of decision.

Takeaways

Takeaways:

1. Consumer Reports is actively developing prototypes to demonstrate how AI can enhance shopping with trusted information

2. The “Buy It For Me” prototype enables complete purchase flows through AI agents without consumers visiting retailer sites

3. The “Pay for Trust” prototype shows how CR’s ratings can be licensed to other AI agents to inform consumer decisions

4. Implementing agentic commerce requires new frameworks for consent, authorization, and experience design

5. CR is collaborating with industry partners and academic institutions to develop ethical AI shopping systems that prioritize consumer interests

Decentralized Governance Models for Urban Sustainability Initiatives.

• Date: 2025-10-25

• Author: Sustainability Directory

Description

Decentralized governance reframes the city not as a machine to be controlled, but as an ecosystem to be cultivated through distributed intelligence and shared stewardship. → Scenario

Summary

Theoretically, a blockchain-based platform could enable transparent, tamper-proof voting on local projects, automate the distribution of green energy credits from a neighborhood microgrid, or manage the property titles of a community land trust with unimpeachable security. These systems offer a mechanism for creating ‘smart contracts’ for sustainability → automated agreements that execute when certain conditions are met, such as a community achieving a recycling target to unlock municipal funds. This includes its ability to integrate with formal legal frameworks, respect informal cultural norms, empower existing community structures, and translate its digital processes into legible, meaningful actions in the physical world.

Takeaways

Takeaways:

1. Decentralized governance offers alternatives to top-down urban planning but requires careful design to avoid creating new inequities.

2. Technology alone cannot solve governance challenges; social, political, and economic systems are equally important.

3. The ‘sovereignty paradox’ warns that tools meant to distribute power can create new forms of centralization and exclusion.

4. Critical pivot points include ownership of digital infrastructure, definition of stakeholders, mandate for interoperability, and investment in human capacity.

5. ‘Social interoperability’ - the ability of governance systems to interface with existing social structures - may be more important than technological efficiency.

6. A municipal protocol with requirements for equity, accessibility, cross-scale linkage, and data stewardship can help steer innovation toward more equitable outcomes.

7. The future of cities depends on whether we view them as platforms for individual gain or commons for collective well-being.

not much happened today

• Date: 2025-10-24

• Author: TBA

Description

vLLM announced support for NVIDIA Nemotron Nano 2, featuring a hybrid Transformer–Mamba design and tunable “thinking budget” enabling up to 6× faster token generation. Mistral AI Studio launched a production platform for agents with deep observability. Baseten reported high throughput (650 TPS) for GPT-OSS 120B on NVIDIA hardware. Hugging Face InspectAI added inference provider integration for cross-provider evaluation. Thinking Machines Tinker abstracts distributed fine-tuning for open-weight LLMs like Qwen3 and Llama 3. In China, MiniMax M2 shows competitive performance with top models and is optimized for agents and coding, while Zhipu GLM-4.6-Air focuses on reliability and scaling for coding tasks. Rumors suggest Gemini 2.5 Flash may be a >500B parameter MoE model, and a possible GPT-5.1 mini reference appeared. Outside LLMs, Tahoe-x1 (3B) foundation model achieved SOTA in cancer cell biology benchmarks. Research from Stanford introduces a method to detect model provenance via training-order “palimpsest” with strong statistical guarantees.

Takeaways

Takeaways:

1. The AI field continues rapid evolution with regular model releases and capability improvements

2. Efficiency and performance optimization remain key focus areas, with models becoming more capable while requiring fewer resources

3. Agent systems are becoming more sophisticated with better memory and orchestration capabilities

4. Multimodal AI is advancing rapidly, particularly in vision-language models and video generation

5. The developer ecosystem is maturing with better tools for building and deploying AI systems

6. Community discussions highlight both excitement about new capabilities and concerns about implementation challenges, costs, and ethical considerations

Reflection: Why we created Common Good AI and why it is so important

• Date: 2024-10-08

• Author: Tom Kehler

Description

The Common Good AI story started at a meeting at MIT Media Lab, orchestrated by John Clippinger in March 2023. We had a shared vision of AI that looked beyond the current fascination with large language models. We gathered a group that included John Cordier, CEO of Epistemix, and Kim Polese (a co-fo

Summary

The time has come to turn attention to AI as fundamental to the development of natural science, tapping into new models on the horizon that included work out of neuroscience and computational biology—the Free Energy Principle and Active Inference. From an application standpoint, we at CrowdSmart have been focused on a new generation of AI technology that is more in line with the adaptive learning style of the neuroscience models of the Free Energy Principle. For that reason, we saw the need to form a non-profit organization that could use the technology to address issues related to social impact, healthcare, education, the development of stronger cities, and deliberative democracy.

Takeaways

Takeaways:

1. Common Good AI was founded in March 2023 with a mission to create AI that serves humanity’s collective needs

2. The organization emphasizes looking to natural science for understanding intelligence rather than just historical data

3. Their approach incorporates the Free Energy Principle and insights from physics about cooperative phenomena

4. Common Good AI technology aims to significantly reduce energy consumption compared to current AI systems

5. The organization will focus on applications that benefit society, including deliberative democracy and social impact projects

6. They are forming collaborations with prestigious academic institutions including MIT, Harvard, and Stanford

7. Common Good AI plans to make their technology openly available for non-commercial use

8. The organization is separate from CrowdSmart, which provides technology support at reduced cost

The Embedded Economy

• Date: 2025-10-24

• Author: Joel John; Sumanth Neppalli

Description

How Prediction Markets Will Change The Web

Summary

They build the venue, then spend the rest of their energy dragging attention back to it—screenshotted odds on X, quote-tweets from power users, Discord bots piping prices into chats. Discovery happens off-platform, intent dies in the tab hop, and the product starts optimising for virality (“markets that screenshot well”) instead of truth (tight rules, fast settlement). Creators trade outrage for calibration; a visible track record becomes a balance sheet they can carry across posts, and rev-share on embedded flow beats chasing the next viral rant.

Takeaways

Takeaways:

1. The internet’s core economy is shifting from capturing attention to enabling frictionless transactions via prediction markets.

2. Prediction markets, particularly when embedded within existing platforms like Twitter/X, allow direct monetization of content and opinions, bypassing traditional ad models.

3. AI agents are crucial for providing liquidity, aggregating information, and executing trades in prediction markets, enabling scalability and efficiency.

4. Embedded prediction markets enhance media credibility and usefulness by turning claims into live, falsifiable probabilities with clear resolution.

5. While prediction markets offer a path forward, they create tension: monetizing everything risks undermining authentic art and expression, suggesting some content will inherently resist commercialization.

6. The explosive growth of prediction markets (e.g., volume surge from 20Mto20Mto2B weekly) signals a major shift in how the web values information and truth.

7. Stablecoins and composable financial primitives are foundational enablers for this transaction-based web evolution.

8. The future web will blur lines between exchanges and content platforms, with embedded markets becoming a key monetization model.

A Tale of Two Realities: The Dual Life of an AI Governance Professional

• Date: 2025-10-13

• Author: Legal; Kris J

Description

AI governance professionals today live in a constant state of duality.

Summary

The contrast is stark - while our day jobs demand precision and predictability, our scroll feeds serve up a kaleidoscope of generative chaos that defies conventional legal boundaries and challenges every governance instinct we’ve been trained to trust. Through the lens of synthetic media, we’ll briefly examine considerations for building robust AI governance programs that are agile enough to respond to the rapid, real-time evolution of generative technologies. The frameworks we design must not only meet legal thresholds but also anticipate the ethical, reputational, and societal ripple effects of AI systems deployed into a world where synthetic media, generative chaos, and blurred realities are becoming the norm.

Takeaways

Takeaways:

1. AI governance professionals must balance structured compliance with responsiveness to rapidly evolving generative technologies.

2. Synthetic media like Sora 2 exposes limitations in current legal frameworks regarding copyright, consent, and identity.

3. Rather than avoiding generative chaos, governance professionals should use it as a living laboratory to test assumptions and identify edge cases.

4. Effective governance requires both technical fluency and adaptive mindset, strategic empathy, and comfort with ambiguity.

5. The same laws and ethical principles may be interpreted differently across enterprise-grade AI and generative expression contexts.

6. Future governance frameworks must be culturally responsive and legally resilient to address the challenges posed by generative technologies.

AI Doesn’t Paint: Why We Cannot Separate Art from the Artist

• Date: 2025-06-02

• Author: Ava Haidar; Nandini Jiva

Summary

As Labonie Roy, an India-based designer and illustrator, noted to us about the Ghibli tradition: “The hopes and imaginations of a post-war culture, the changing landscape of the countryside, the legends and spirits that are interwoven into people’s lives. Livelihood and ownership concerns will always be relevant to the question of AI-generated art, but tend to reduce the policy scope of what is at stake to economic exclusions and the legal dealings of Big Tech companies. There is thus a need to reimagine the grounds for AI-art regulation to include vocabulary that touches upon not just fair use and the legal entitlement to any artwork, but also to reasonably accommodate the bottom line of protecting artistic and creative work, to preserve our agency, and mitigate the loss of our capacity as humans.

Takeaways

Takeaways:

1. AI-generated art raises concerns beyond legal copyright issues, including the loss of human intent, connection, and effort in creative processes.

2. The ‘Ghiblification’ trend highlighted how AI can strip away the nuanced, emotive details that make art meaningful to human audiences.

3. Current copyright frameworks may be insufficient to address the unique challenges posed by AI-generated art, particularly regarding moral rights.

4. Artists like Hayao Miyazaki view AI-generated art as fundamentally antithetical to the human essence of creativity.

5. Regulation of AI in art should consider not just economic impacts but also the preservation of human agency and the intrinsic value of artistic creation.

6. The debate over AI in art ultimately raises profound questions about what constitutes art and whether it can meaningfully exist without human personhood.

Saving democracy for the price of a swimming pool

• Date: 2025-10-21

• Author: Open Indie

Description

Ten walking-minutes from where I live in Oslo, Norway, there’s a large pool that opened at the start of this year. I love this pool; I wa...

Summary

These services must champion algorithmic pluralism, ban recommendation engines that reward outrage and hate, and give users explicit control over what they see, no more “revealed preferences” that mask surveillance. Bluesky’s continued liberties are contingent on an ever-closer collaboration with European nations that can act as a backstop against authoritarian censorship by providing an offramp to digital dissidents and a parallel polis for online discourse at large. What I hope to make obvious with my swimming pool analogy is that this is a trivial amount of money for a country of Norway’s immense wealth, and every day we fail to effectively deploy our capital is a strategic failure of historical proportions.

Takeaways

Takeaways:

1. Europe’s reliance on US digital infrastructure threatens democratic values and sovereignty

2. Investment in European digital platforms is a strategic necessity, not just an economic opportunity

3. Open-source protocols like Bluesky offer a viable alternative to billionaire-controlled platforms

4. Norway’s wealth positions it to make significant impact in the digital sovereignty movement

5. The $200M cost of a swimming pool represents a trivial amount for Norway’s strategic interests

6. Without proactive investment, Norway risks becoming irrelevant in the digital landscape

State of Crypto 2025: The year crypto went mainstream

• Date: 2025-10-22

• Author: Daren Matsuoka; Robert Hackett; Jeremy Zhang; Stephanie Zinn; Eddy Lazzarin

Description

This year’s report explores the data behind crypto’s maturation, from institutional adoption and the rise of stablecoins to crypto x AI.

Summary

The shift from a hostile regulatory environment to a much more supportive one, alongside accelerating adoption of these technologies — from stablecoins to the tokenization of traditional financial assets to other emerging use cases — will define the next cycle. This trend appears to be cooling down in recent months — there were 56% fewer launches in September than in January — as sound policy and bipartisan legislation clears the way for more productive blockchain use cases. Protocol standards such as x402 are emerging as a potential financial backbone for autonomous AI agents, helping them make micro-transactions, access APIs, and settle payments without intermediaries — an economy that Gartner estimates could reach $30 trillion by 2030.

---

Join the curation team

If you like the Distroid Digest and you are interested in curating media on the human-technology relationship for yourself or others, consider sending us relevant media whenever you have the time so it can appear in the Distroid Digest and be catalogued in the Distroid Catalogue Knowledge Graph (DCKG).

For more information on how to help out and join our team, contact us at ledgerback@gmail.com.

Donate

Please consider donating to Ledgerback to support the work on the Distroid project.