.svg)
Feedback about a Fixer Contact
Knowledge Base
This section offers a curated selection of up-to-date materials and research on the use of artificial intelligence in the media and beyond.
AI Regulation in the EU
The document from the European Commission outlining its approach to AI regulation. It describes risk classification, transparency requirements for generative models, and principles for user protection.
The State of AI 2024
The annual global report by the international management consulting firm McKinsey which examines the state of AI development. The authors analyze technology adoption rates across industries, corporate investment levels in AI and the sectors where generative models are most prevalent. The material provides insight into strategic trends and anticipated labor-market shifts.
What is Generative AI and How Does it Work?
The article from the University of Arizona explains how AI analyzes data and transforms it into new forms — from music and design to potential medicine discovery. The authors outline the main stages of generative system operation, their labor-market impact, and implications for future professionals.
What is Generative AI? Understanding How GenAI Works
The explainer from Dartmouth College clarifying the operating principles of large language models. The authors stress that generative AI is not a search engine — it does not “find” information but generates it based on statistical patterns in data. The piece helps readers understand the limits of AI use in learning.
What is Generative AI?
The guide from the University of Michigan for faculty and students offers a list of AI tools suitable for educational use. It also includes a glossary of key terms to help readers navigate AI-related concepts quickly.
Machine learning and generative AI: What are they good for
The report by MIT Sloan researchers explains the distinction between machine learning and generative AI. The authors show that classical machine learning excels at predicting or classifying data (e.g., in finance or logistics), while generative AI is effective in creating text, visual content, and innovative solutions. The material clarifies when to apply each approach.
What is Generative AI?
The article prepared by researchers from the University of Pittsburgh explains how generative AI works — what it means to create content “from scratch,” and which models (ChatGPT, Bard, Copilot) are currently the most popular. The authors describe how this technology can assist students and educators but also warn about risks such as academic dishonesty and the need for critical evaluation of AI outputs.
Traditional AI vs. Generative AI: What's the Difference?
The publication from the University of Illinois outlines the differences between traditional artificial intelligence approaches and the new wave of generative models. The authors note that classical AI focused on data analysis and decision-making in narrow fields, while generative models can create new text, images, or audio.
Generative AI and The Guardian: ‘What we do can’t be reproduced synthetically’
The article explains The Guardian’s cautious stance toward generative AI. The author describes why the newsroom is postponing major decisions, what internal debates are taking place around rights, attribution, and risks, and how these factors influence external partnerships and policies.
BBC Debuts Generative AI Tools: Transforming News Accessibility and Style
The news piece details how the BBC is implementing generative AI in its news production through two key tools. “At a Glance” creates concise, list-style summaries of articles, a format especially relevant to younger audiences who prefer quick content. “BBC Style Assist” reformats local stories to match the corporation’s editorial tone, using a large language model trained on the BBC’s own article corpus.
The Washington Post introduced AI bots for readers last year. Here’s how it’s going.
The article in Technical.ly reports on the practical performance of The Washington Post’s AI tool “Ask The Post AI,” launched in November 2024. The author notes that the tool makes mistakes — for example, it incorrectly named Joe Biden as the current president instead of Trump and listed Lenin among participants in the Cold War. The outlet acknowledges the experimental nature of the tool and urges readers to verify information, as “AI can be wrong.”
Washington Post, Virginia Tech collaborate on AI news search tool
The article describes the collaboration between The Washington Post and Virginia Tech, which developed an AI tool for searching the newspaper’s archives. Unlike typical large language models, this tool uses retrieval-augmented generation: it first searches for relevant articles exclusively within The Washington Post archive and then generates summaries based solely on that content, avoiding “hallucinations.” The model refuses to answer if it lacks sufficient information from credible sources.
