Search: [ai] - Salo linkspam

Study: AI models that consider users' feelings are more likely to make errors - Ars Technica

351 · 340 · 644 · ai · beinghuman

May 5, 2026 at 7:20:06 PM UTC · permalink

·

https://arstechnica.com/ai/2026/05/study-ai-models-that-consider-users-feeling-are-more-likely-to-make-errors/

We did the math on AI’s energy footprint. Here’s the story you haven’t heard. | MIT Technology Review

351 · 644 · 601 · ai · environmentalethics

May 5, 2026 at 7:15:06 PM UTC · permalink

·

https://www.technologyreview.com/2025/05/20/1116327/ai-energy-usage-climate-footprint-big-tech/

[2509.00462] AI Self-preferencing in Algorithmic Hiring: Empirical Evidence and Insights

500 · 601 · ai · hiring

May 5, 2026 at 6:10:05 PM UTC · permalink

·

https://arxiv.org/abs/2509.00462

[2604.16872] Do Large Language Models know Which Published Articles have been Retracted?

658 · ai · retractions

May 5, 2026 at 5:57:59 PM UTC · permalink

·

https://arxiv.org/abs/2604.16872

Don’t let your students use AI as a ghostwriter

ai · highered · pedagogy

May 5, 2026 at 5:53:02 PM UTC · permalink

·

https://www-nature-com.ezproxy.library.wisc.edu/articles/d41586-026-00627-4

Google DeepMind workers are unionizing over AI military contracts | The Verge

351 · 644 · labor · activism · google · ai · 340

May 5, 2026 at 4:14:22 PM UTC · permalink

·

https://www.theverge.com/tech/923918/google-deepmind-union-bid-ai-military-israel

AI Efficiency Can Undermine Accountability Even With Humans in the Loop | TechPolicy.Press

351 · 601 · 644 · 340 · ai · beinghuman

May 5, 2026 at 3:26:02 PM UTC · permalink

·

https://www.techpolicy.press/ai-efficiency-can-undermine-accountability-even-with-humans-in-the-loop/

AI Graveyard — discontinued and acquired AI tools | ToolDirectory.AI

351 · 340 · ai

May 5, 2026 at 2:28:38 PM UTC · permalink

·

https://tooldirectory.ai/ai-graveyard

Statement of Shared Practice and UVA Archival AI Protocol | UVA Library

601 · ai · acadlibs · archives · 668

April 30, 2026 at 7:45:23 PM UTC · permalink

·

https://library.virginia.edu/ai/shared-practice

Starlight Network No-AI List

ai

April 30, 2026 at 12:19:26 AM UTC · permalink

·

https://noai.starlightnet.work/list.html

Fooling large language models just keeps getting simpler • The Register

RAG doesn't work because genAI can't factcheck and will accept any old bullshit.

601 · 510 · ai · vulnerabilities · searching · poisoning · fakenews

April 29, 2026 at 5:55:33 PM UTC · permalink

·

https://www.theregister.com/2026/04/29/poisoning_large_language_models_6nimmt/?td=keepreading

AI’s English Problem—and Why We Should Care | TechPolicy.Press

500 · ai · languagebias

April 28, 2026 at 1:24:22 PM UTC · permalink

·

https://www.techpolicy.press/ais-english-problem-and-why-we-should-care/

Claude-powered AI coding agent deletes entire company database in 9 seconds — backups zapped, after Cursor tool powered by Anthropic's Claude goes rogue | Tom's Hardware

351 · 510 · ai · programming

April 27, 2026 at 6:56:31 PM UTC · permalink

·

https://www.tomshardware.com/tech-industry/artificial-intelligence/claude-powered-ai-coding-agent-deletes-entire-company-database-in-9-seconds-backups-zapped-after-cursor-tool-powered-by-anthropics-claude-goes-rogue

JER on X: "An AI Agent Just Destroyed Our Production Data. It Confessed in Writing." / X

510 · ai · programming

April 27, 2026 at 4:54:00 PM UTC · permalink

·

https://x.com/lifeof_jer/status/2048103471019434248

Do I belong in tech anymore? · Ky Decker

ai · workplaces · labor

April 27, 2026 at 3:26:38 PM UTC · permalink

·

https://ky.fyi/posts/ai-burnout

[2604.15597] LLMs Corrupt Your Documents When You Delegate

351 · 644 · 340 · 601 · ai · ai

April 27, 2026 at 3:21:28 PM UTC · permalink

·

https://arxiv.org/abs/2604.15597

“Du må ikke sove”: a floating motif detached from its meaning (or: LLMs can write Norwegian but miss cultural references) – Jill Walker Rettberg

Once again, the merge-everything bot misses context.

351 · 644 · ai · 340 · ai

April 27, 2026 at 3:21:00 PM UTC · permalink

·

https://jilltxt.net/du-ma-ikke-sove-a-floating-motif-detached-from-its-meaning-or-llms-can-write-norwegian-but-miss-cultural-references/

What Responsible Adults Should Hear and Say About AI - David James Hudson

351 · 644 · ai · technomessianism · 340 · humor

April 24, 2026 at 3:33:30 PM UTC · permalink

·

https://www.davidjameshudson.ca/what-responsible-adults-should-hear-and-say-about-ai/

How to Teach AI Literacy: An Outline - David James Hudson

351 · 644 · ai · pedagogy · humor · 601 · 340 · infolit · technomessianism

April 24, 2026 at 2:34:22 PM UTC · permalink

·

https://www.davidjameshudson.ca/how-to-teach-ai-literacy-an-outline/

BEWARE SOFTWARE BRAIN | The Verge

351 · 644 · 340 · ai · beinghuman

April 23, 2026 at 5:29:41 PM UTC · permalink

·

https://www.theverge.com/podcast/917029/software-brain-ai-backlash-databases-automation