Welcome to "From the Research Desk"

Welcome to the first edition of “From the Research Desk!” This periodic communication aims to share some of what I and the Siegel Family Endowment’s research team have been mulling over, paying attention to, grappling with, and collaborating on. As an organization, Siegel aims to “understand and shape the impact of technology on society” and recognizes that the process of understanding is not linear; no single research project or experiment will bring full understanding of the complex systems that sit at the intersection of technology and society, if that were even possible. Rather, it is the act of swimming around in the primordial pool of information, research findings, individual expertise, common sense, inquisitive dialogue, and introspection that can point us towards greater understanding.

In this newsletter, I will aim to share half-thoughts, interesting readings, and emerging ideas with the hope that they might prove to be of interest to others. I will be more likely to ask questions and less likely to offer explanations. The content will likely be a mix of the obvious, the obscure, and the complex.

At my perch leading a foundation-based research team, I have the privilege to speak and collaborate with a wide range of nonprofit leaders and researchers across sectors. This letter series aims to do the same—picking up threads of ideas from many sources, silos, and disciplines. Sometimes, what appears to be an isolated or unrelated insight can be the flip side of a novel idea elsewhere.

I hope you find this newsletter to be a useful addition to your information diet!

Question(s) of the Moment:

Are data centers constructed by private companies core digital “infrastructure,” or are they noisy, polluting eyesores that represent growing economic inequality (or both)? And, for that matter, what is infrastructure anyway? (See below!)

Factoid of the Moment:

“..nine out of 10 teachers reported that their schools use student activity monitoring software, including 39% that monitor outside of school hours, and 29% that monitor personal (e.g. non-school-issued) devices.”

Via Testimony of Alexandra Reeve Givens, President & CEO, Center for Democracy & Technology before the U.S. House Committee on Education and Workforce, January 14, 2026 citing Elizabeth Laird, Maddy Dwyer & Hannah Quay-de la Vallee, “Hand in Hand: Schools’ Embrace of AI Connected to Increased Risks to Students,” Center for Democracy & Technology, Oct. 8, 2025

Top of Stack

Generative AI continues to dominate discussions related to the future of work, workers, and the workplace. Much of the discussion focuses on the impact on an individual employee — will AI displace jobs or work? How will people use AI in the workplace to augment their skills? However, there is much less research or thought about AI’s role in collaborations with groups of people and what collaboration looks like in that context. A recent paper explores this idea of “Generative Collective Intelligence” and is worth a read.

“Amplifying Human Creativity and Problem Solving with AI Through Generative Collective Intelligence” by Thomas P. Kehler, Scott E. Page, Alex Pentland, Martin Reeves, John Seely Brown:

“In reframing AI as generative collective intelligence (GCI), we place primary emphasis on bridging cognition between humans and AI. We do not deny the awesome abilities of AI acting alone, nor do we deny that it improves by the minute. But it is our belief that the greatest potential of AI rests in GCI, in collaborations that combine human creativity and wisdom with AI’s computational and organizational capabilities.”

Science Fiction predicts Future (again)

The Washington Post reports that the AI company Anthropic ran a program in 2024 called “Project Panama” that aimed to “destructively scan all the books in the world” by acquiring books, slicing off their spines, scanning them, and then recycling them.

Fans of Vernor Vinge will surely recognize this as a plot line from Rainbows End. In the novel, the Librareome Project simultaneously shreds and scans books from a local library, with plans to make them available exclusively in digital form. Not only do the projects largely line up, but so too does the timeline; while Rainbows End was published nearly two decades ago, the events in the novel take place in 2025!

See WaPo’s Inside a tech company’s secretive plan to destroy millions of books.

Stay in Touch

Follow/connect with me on bluesky at https://bsky.app/profile/argmax.com or on LinkedIn at https://www.linkedin.com/in/john-irons/

Share From the Research Desk

Deeper Dive: What is Infrastructure?

With a big HT to Resident Fellow Ingrid Burrington who recently shared this 2021 article from Justinien Tribillon: “Inventing ‘infrastructure’: tracing the etymological blueprint of an omnipresent sociotechnical metaphor”

“‘Infrastructure’ is a popular expression because it is evocative: it encapsulates modernity—the association of capital, statal project, and technology. Arguably, we collectively share a networked imaginary where our social and economic structures rest on a series of interdependent systems that enable them to function: the roads and rail tracks that bring us to work in the morning, the internet and mobile networks that allow us to watch Netflix on our laptops, our power-grids and water pipes.”

Interestingly, as Tribillon tracks down, the word infrastructure comes from 19th century French, with the initial appearance in a legal document founding a Russian Rail system:

“The legal document signed between Alexandre II of Russia and the consortium led by the Brothers Pereire, that marked the foundation of the Great Company of Russian Railways in January 1857 used for the first time the term infrastructure and the dualism infrastructure/superstructure.”

The early french use of the word also included the source of financing as integral to the definition:

“This dualism between what is below the railway and funded by the State, and what sits above and financed by private actors established in the law of 1842 now had acquired names of their own: infrastructure and superstructure….Infrastructure is the invention of financiers invested by an industrialist utopia, not that of engineers.”

Codified in French dictionaries, then later making its way to English (unaltered from the French), the term ranged from the physical underpinnings and literal foundations of railways and later buildings and other structures. Though a detour through French Marxism apparently happened as well:

“Yet from the late 1890s onwards and throughout most of the 20th century, the denomination infrastructure vs superstructure became a convention of French Marxist semantics…”

In the 20th century, the term has expanded, covering multiple concept, contexts, and disciplines – sometimes physical, but more often metaphorical – and which, Tribillon argues, is still consistent with its origins:

“The idea of a spatial, or metaphysical dualism between what is below and above did not start with the concept of infrastructure—it links ancient religions to postmodern philosophies. Yet, infrastructure as an expression of modernity crystallised a sociotechnical imaginary, a relation between technology, space and power.

—

Additional Reading suggested from AI (Claude)
*Summaries may be inaccurate, and I do not vouch for the value of content!*

Student Monitoring & EdTech Surveillance

• Digital Dystopia: The Danger in Buying What the EdTech Surveillance Industry is Selling — American Civil Liberties Union (ACLU), 2023. A comprehensive examination of the K-12 EdTech surveillance industry, its unsubstantiated safety claims, and the harms these products cause to students, particularly those from marginalized communities.

• The Surveilled Student — Danielle Keats Citron, Stanford Law Review, Vol. 76 (2024). A detailed legal analysis of student surveillance practices, arguing that continuous monitoring harms students’ intimate privacy, freedom of expression, and civic development.

• Public Schools, Private Eyes: How EdTech Monitoring Is Reshaping Public Schools — New America Foundation, 2025. An overview of how AI-powered surveillance tools are expanding in K–12 settings, with particular attention to risks for undocumented students and other vulnerable groups.

• School-Based Online Surveillance of Youth: Systematic Search and Content Analysis of Surveillance Company Websites — O’Daffer et al., Journal of Medical Internet Research, 2025. The first systematic study of companies offering school-based online surveillance services, examining their monitoring capabilities and data collection practices.

AI, Work, and Organizational Productivity

• Artificial Intelligence and the Future of Work — National Academies of Sciences, Engineering, and Medicine, 2025. A comprehensive assessment of AI’s implications for economic productivity, workforce displacement, and education, with policy recommendations for workers and policymakers.

• Generative AI, Productivity and the Future of Work — Federal Reserve Bank of St. Louis, 2025. Research finding that generative AI users saved an average of 5.4% of work hours weekly by mid-2024, with adoption rates tracking early personal computer diffusion patterns.

• The Projected Impact of Generative AI on Future Productivity Growth — Penn Wharton Budget Model, 2025. An economic modeling study estimating that around 42% of current jobs are potentially exposed to AI automation, with productivity contributions expected to grow substantially over the next decade.

• New Future of Work Report 2025 — Microsoft Research, 2025. A cross-disciplinary review of AI’s effect on workplace collaboration, arguing the next frontier is collective rather than individual productivity—how teams and organizations can leverage AI together.

Data Centers, Digital Infrastructure, and the Public Interest

• Regulating the Digital Giants: Policy and Public Responses to the Data Center Boom — Yale Clean Energy Forum, 2025. Examines how data centers are increasingly being treated as public-interest infrastructure subject to environmental and social accountability, with comparisons of U.S., European, and local regulatory approaches.

• Data Center Energy Infrastructure: Federal Permit Requirements — Congressional Research Service, 2025. A nonpartisan overview of federal policy and permitting requirements for data centers, including electricity consumption projections and the evolving regulatory landscape under recent executive orders.

• The Making of Critical Data Center Studies — Edwards, Cooper, & Hogan, Convergence: The International Journal of Research into New Media Technologies, 2024. A survey of interdisciplinary scholarship examining data centers not merely as technical infrastructure but as sites revealing the politics, ecologies, and power relations of the digital economy.

Copyright, AI Training Data, and the Knowledge Commons

• Copyright and Artificial Intelligence — U.S. Copyright Office, 2024–2025. A three-part report series addressing digital replicas, the copyrightability of AI-generated works, and—most relevant here—whether using copyrighted works to train generative AI constitutes fair use.

• Copyright and AI Training Data—Transparency to the Rescue? — Journal of Intellectual Property Law & Practice, Oxford Academic, 2025. Analyzes the growing calls for mandatory disclosure of AI training datasets from rights holders and governments, and the legal and policy arguments on all sides.

—

LLM statement: GAI tools (Gemini, Claude) were used to lightly copyedit this article, to suggest additional embedded links, and to generate the additional reading section.