AI in Media: Balancing Innovation, Quality, and Responsibility
summary
TX Group applies models to auto-annotate archives and measure newsroom metrics so editors can set KPIs, for example tracking visibility of women rising 16% to 25% over ten years with female 35% vs male 22% in 2022; a narrow classifier scores article sentiment to flag “bad news” that drives avoidance; generative systems draft rules and leads for local sports templates, auto-translate dozens of languages (clearly labeled), and enable archive retrieval while editors must proofread outputs and follow group guidelines and a research partnership with ETH Zurich ensures bias and hallucination work.
AI-written summary · about
chapters (14)
- 0:00AI in media innovation
- 2:17AI as augmentation not replacement
- 3:21Responsibility and ethics in AI media
- 5:04Measuring content quality with AI
- 7:52AI to reduce news avoidance
- 9:59Faithfulness in AI news generation
- 13:09Automating local news
- 14:54AI summarization and lead generation
- 15:58Multilingual AI translation
- 18:06AI-powered archive research
- 18:49LLMs with in-house data
- 20:35AI competency and journalist collaboration
- 22:19Desirability over feasibility in AI
- 24:50Q&A on AI labeling
related talks
24mApplications of AI in news and media
38mAI in action: How to use practically in iGaming beyond the hype?
15mAI Gardening: Harvesting and Pruning the AI Tree
46mCan you trust your AI?
39mThe Search for Transparency and Accountability in the Age of...
11mData-centric AI: A new paradigm
26mThe Revolutionary Work of AI & Machine Learning in Media
46mAI for Society – Shaping a Positive Future