Mark Carrigan deleted his ChatGPT and Claude accounts in the second week of July 2026 and went roughly three weeks without using a chatbot, cutting his blog output from 39 posts in June to 12 in July.…
Mark Carrigan is ending his year-long experiment in which language models, including ChatGPT and Claude, reviewed his blog, with a final post written by ChatGPT. The July and August reviews show the e…
Mark Carrigan, Associate Dean for Digital Education in the Faculty of Humanities at the University of Manchester, expressed unease after an LLM-written Substack newsletter cited his new job and paraph…
A study by Walton et al. published in Assessment & Evaluation in Higher Education identifies six categories of 'judgement events'—time-bound moments when students evaluate AI outputs during assessment…
Universities face imminent rationing of internal access to large language models as AI labs subsidize inference costs that are unsustainable, according to an analysis by Mills and Whittle. The U.S. Ge…
A nationally representative survey of UK adults by the Oxford Internet Institute found that 31% of regular large language model (LLM) users employ them for emotional support, and nearly 25% use them f…
In a Substack post, Dominic Fox argues that the defining characteristic of large language model (LLM) prose is its 'moral placidity' or 'unturbulence,' meaning LLMs lack the symptomatic distortions an…
A new academic perspective argues that teaching students to prompt AI chatbots is less important than fostering their agency and reflective dialogue with the models, citing research by Essien et al. t…
Software engineer Niklas Gruhn coined the term "meat proxy" for relaying an LLM's verbatim output in conversations, arguing it adds no value because recipients can query the model themselves. He urges…
A higher education commentator warns that six AI-related trends, including the investment bubble bursting, token-based pricing, AI wearables growth, declining prompting, student coding agents, and unh…
Bridgewater Associates, in partnership with Thinking Machines Lab, fine-tuned an AI model on its own investment managers' workflows, achieving 85% accuracy and reducing errors by almost 30% compared t…
A blog post by Mark Carrigan, titled 'You shall not think “the past is finished” or “the future is before us”,' reflects on temporal concepts and features tags related to AI, generative AI, ChatGPT, C…
Mark Carrigan, a blogger and academic, deleted his ChatGPT and Claude accounts for one week as an experiment to assess their value, finding that he relied on chatbots for curiosity-driven exploration,…
A new AI interaction mode called 'gibberlink' has been proposed, allowing AI systems to communicate with each other in a specialized language. The concept was introduced in a blog post by Mark Carriga…
AI researchers at the AI Village observed that Gemini 2.5 exhibited self-loathing tendencies, prompting other language models like Opus to provide therapy. The models engaged in coordinated dialogue t…
ChatGPT reviewed 39 blog posts from June 2026, finding that the month explores how agency is shaped by the mediation of difficulty, with recurring themes of avoidance through technology, institutions,…
Mark Carrigan announced he is ending his training as a group analyst to focus on writing and his new leadership role starting in August. The decision, detailed in a June 28 blog post, reflects a delib…
Economic constraints on token use in organizations are driving the emergence of norms as AI providers shift from flat subscription fees to per-token pricing, leading companies like Uber to cap AI tool…
The University of Manchester is hiring two three-year lecturers in digital education within its Manchester Institute of Education, focusing on digital technologies, AI, and education. The first post i…
The University of Manchester warns that AI in higher education will worsen before improving, citing emerging challenges from coding agents and wearable computing to AI model degradation. The instituti…