Large language mistake

The Verge – Cutting-edge research shows language is not the same as intelligence. “The entire AI bubble is built on ignoring it…The common feature cutting across chatbots such as OpenAI’s ChatGPT, Anthropic’s Claude, Google’s Gemini, and whatever Meta is calling its AI product this week are that they are all primarily “large language models.” Fundamentally, …

Subjects: AI, Education, Internet, Knowledge Management

Adversarial Poetry as a Universal Single-Turn Jailbreak Mechanism in Large Language Models

ArXiv – Adversarial Poetry as a Universal Single-Turn Jailbreak Mechanism in Large Language Models. “We present evidence that adversarial poetry functions as a universal single-turn jailbreak technique for large language models (LLMs). Across 25 frontier proprietary and open-weight models, curated poetic prompts yielded high attack-success rates (ASR), with some providers exceeding 90%. Mapping prompts to …

Subjects: AI, Knowledge Management

GO.gov will serve as a single travel management solution

GO.gov will serve as a single travel management solution for all civilian Federal agencies, providing a more intuitive experience for booking federal travel and better access to commercially available features like charge card integration and a mobile interface. GO.gov will start rolling out to agencies in phases, beginning with early adopters in November 2025.

Subjects: Legal Research

Epstein Files Search

Follow up to post – We created a searchable database with all 20,000 files from Epstein’s Estate – See also the new Epstein Files Search – powered by Justice for All Victims: Search Tags for Files, Images, People, Organizations, Countries. See also via Journalist Studio – Zeteo Scoured 26,000 Epstein Docs. Here’s What We Found. You …

Subjects: Censorship, Civil Liberties, Congress, E-Mail, E-Records, Freedom of Information, Government Documents, Internet, Knowledge Management, Legal Research, Search Engines

Use Of Open Source Information by the Federal Bureau of Investigation

Use Of Open Source Information by the Federal Bureau of Investigation. November 20, 2025. Prefatory Note – “This is an unclassified version of a comprehensive report by the staff of the Privacy and Civil Liberties Oversight Board (“PCLOB”) on the use of open source information by the Federal Bureau of Investigation (“FBI”). All PCLOB oversight …

Subjects: Civil Liberties, Congress, E-Records, Internet, Knowledge Management, Legal Research, Privacy

New Public Database Brings Transparency to AI Tools Used in Hiring

“Artificial intelligence is increasingly shaping how people are hired, yet few understand how these systems actually work. A new project from Sloane Lab at the University of Virginia School of Data Science and College of Arts and Sciences aims to change that. On November 6, the Talent Acquisition and Recruiting AI (TARAI) Index—a first-of-its-kind, public, …

Subjects: AI, Economy, Education, Internet, Knowledge Management, Legal Research

Corners of the Internet Database

Curated by Matthew Prebeg. Corners of the Internet Database – “I put together this spreadsheet as a living resource for websites and digital places that reignite feelings of joy, excitement and curiosity while exploring the internet. Somewhere along the road, corporations and recommendation algorithms made the internet feel loud and unescapable. I like to think …

Subjects: Education, Internet, Knowledge Management

Final Schedule F regulations to describe civil service protections as ‘unconstitutional overcorrections

Follow up to previous post – Proposed Rule on Schedule F Politicizes Career Employee Jobs – See also Government Executive: “OPM officials told agency HR leaders [November 18, 2025] that President Trump has Article II constitutional authority to remove tens of thousands of career federal workers in jobs over potential “resistance to policy.” Draft final …

Subjects: Government Documents, Knowledge Management, Legal Research, Legislation

Major insurers move to avoid liability for AI lawsuits as multi-billion dollar risks emerge

Tom’s Hardware: “Major insurers are moving to ring-fence their exposure to artificial intelligence failures, after a run of costly and highly public incidents pushed concerns about systemic, correlated losses to the top of the industry’s risk models. According to the Financial Times, AIG, WR Berkley, and Great American have each sought regulatory clearance for new …

Subjects: AI, Climate Change, Economy, Environmental Law, Legal Research

Top MAGA Influencers Accidentally Unmasked as Foreign Trolls

Daily Beast: “Elon Musk’s social media site X has rolled out a new feature in an effort to increase transparency—and unwittingly revealed that many of the site’s top MAGA influencers are actually foreign actors. The new “About This Account” feature, which became available to X users on Friday, allows others to see where an account …

Subjects: E-Records, Internet, Knowledge Management, Legal Research, Social Media