Stephen C. Webster

"Last week, SPARC heard from customers that the Copyright Clearance Center’s (#CCC) Get It Now (#GIN) on-demand purchasing service is being discontinued in June. This news follows recent reports that #SpringerNature plans to remove its content from another popular on-demand service, Article Galaxy Scholar (#AGS), by the end of March." https://lnkd.in/eZ3JA8dz #ResourceSharing #DocumentDelivery #ILL #scholarly #libraries #ResearchLibraries #AcademicLibraries

2

The state of Vermont in the USA just passed a law requiring all AI deep fakes to be clearly labelled as such if shared within 90 days of an election. Under the new law, any campaign media used within 90 days of an election that includes AI-generated images, audio, or video must include a disclosure. That disclosure must be visible to the average viewer and large enough to read easily. The legislation also includes rules and restrictions surrounding deepfakes made of candidates. It's just way too easy to reproduce a politician's voice. How is your state handling election-related AI-generated deep fakes? https://lnkd.in/gGQwTttC #AI #Elections #DeepFakes

1

NEW POST: SNJ’s Matt Keer has dug behind DfE research data into #EHCP creation times. He’s found some “deranged” numbers that can’t be real—but does the government really care as long as it gets what it wants? https://lnkd.in/exMy3WZT #SaveOurChildrensRights

4

A notable AI-focused change emerging in the VS Code 1.110 cycle is native browser integration in chat, described as enabling interaction with page elements, screenshot capture for visual debugging, and access to real-time browser console logs directly by AI agents. Other items in the same Insiders cycle include /fork for branching sessions, prompt queueing, background-agent slash commands, and MCP-related updates such as sandboxing for local stdio MCP servers. The VS Code February 2026 iteration plan indicates the stable ship is targeted for early March.

1

SecureWorld News: "By establishing clear guardrails, such as transparency requirements and prohibitions on AI practices that intentionally cause certain harms, where they are clearly warranted, yet not over-regulating in areas that need more development and maturation to better flesh out the issues, Texas is setting strong example for reasonable AI regulation." Thanks to Shawn Tuma for the commentary, as well as Violet Sullivan, CIPP/US CIPM, Ankit G., and KJ Haywood. https://lnkd.in/gRsufAHv

13

1 Comment

Distorted Data and Looming Disaster A recent article titled "Vanishing Archives, Faltering Algorithms" (link in comments) explores how China’s efforts to censor history and control information are undermining its AI progress. With archives erased and narratives tightly managed, the AI models built on that data are producing flawed outputs—and bad decisions. (Check out the hallucination rates!) One key message stands out: AI is only as good as the data it's trained on. It’s critical to work with research partners who prioritize clean data systems and use transparent, purpose-built AI. Relying on generic platforms or opaque black-box tools introduces unnecessary risk. Trust in your inputs defines the value of your insights. #marketingresearch #aiethics #trustyourdata #qualitativeresearch #aiinsights #dataintegrity #insightstrust #mrx

5

3 Comments

Great new report from NISO's Open Discovery Initiative (ODI) about industry perceptions of how GenAI is transforming library discovery There are too many insights to list, but the survey of 236+ libraries & content providers has revealed to me their shared hopes, such as improved content visibility, more accurate recommendations, and time savings for staff & researchers. Respondents' top concerns included that libraries worry about: content fabrication, lack of transparency, AI summary quality, environmental impact of AI data center. Meanwhile publishers have their own concerns such as: copyright infringement, content misrepresentation, and transparency gaps My key report takeaway is that both libraries and publishers want AI benefits but need transparency and trust mechanisms. Read the full report: https://lnkd.in/eDD_-JMb Kudos to National Information Standards Organization and the folks on the ODI like Parvathi (Bobbi) Patham, Teresa Hazen, and the rest committee for such an insightful, important, and timely report.

12

Ahead of our Advanced AI Workshop at SfAA in Albuquerque this week, I've open-sourced more of the tools I use in my own research as part of the AI Anthropology Toolkit. The update includes notebooks I use for my own research — tools for searching academic literature, collecting news and media data, mapping public discourse through search behavior, analyzing discourse structure through text networks, and discovering topics in document collections. All but one run in the browser through Google Colab, most don't need API keys, and all export structured data ready for qualitative analysis.  A few highlights:  - Academic Literature Explorer — searches 250M+ scholarly works via OpenAlex  - Text Network Analysis — co-occurrence networks with community detection for discourse analysis  - Topic Modeling with BERTopic — transformer-based topic discovery with zero-shot mode - Named Entity Recognition with GLiNER2 — zero-shot extraction of people, places, organizations, and  concepts from text  - Google News, Trends, Ngram, Patents, and Scholar explorers for contextual research data  - YouTube Video Search and Transcript Fetcher for media research  - PubMed Literature Harvester for biomedical and health sciences work Stay tuned. There is more to come this week.  GitHub: https://lnkd.in/eKuByA2b #AIAnthropologyToolkit #anthropology #AIanthropology #AI #computationalanthropology #qualitativeresearch #digitalmethods #opensource #SfAA2026 Society for Applied Anthropology Society for Applied Anthropology

370

7 Comments

The Sovereign Architect: Principles and Axioms by Eldan W. Horai (Book Review #2336) #bookreview #NonFiction, #Selfhelp & #Selfimprovement https://lnkd.in/eVeaAs7V via @talesreview

2

I have always been an early adopter. AI is moving faster than any single person or institution can fully process. It operates across layers of abstraction so quickly that by the time we try to describe it, it has already changed. What you are posting right now will drastically evolve by afternoon. Yesterday AI made something visible to me. Not that the FDA is irrelevant, but that it risks becoming late. And in rare disease, late is not neutral. N of one, Undiagnosed, Ultra rare, and Rare Patients and families cannot wait for systems to catch up. We will move on to better models that actually reflect lived experience, heterogeneity, and real outcomes. Those models already exist. They are being built quietly by people who had no choice but to innovate. The question is not whether AI will change medicine. It already has. The question is whether our institutions adapt in time to still matter. Rare disease has been ready for this moment for decades are you? #raredisease #ultrarare #nofone #FDA #HHS #CMS #AI

4