Hello, and welcome to a new edition of The Fix’s monthly special newsletter series where we’re curating new and archival coverage on one important topic.
Polish newspaper Gazeta Wyborcza formulated two rules for its journalists on using ChatGPT:
Otherwise, Wyborcza’s journalists are free to use the tool for supplementary tasks like brainstorming or coming up with interview questions.
This approach seems to be a consensus among news outlets who’ve thought about this topic (including The Fix itself). We all know by now that ChatGPT is bad at generating interesting, engaging, factually correct text from scratch – and extremely effective at creating persuasive but totally incorrect text. We also know that it’s great for tasks like language help and idea generation and best thought of as your assistant and companion rather than a ghostwriter.
But a lot of difficult and interesting questions remain. Is it ethical to use image generation tools like MidJourney to illustrate articles? How to report on fake AI-generated images so as not to amplify their negative effect? What are some non-obvious ways to use text generation tools to make your newsroom more effective?
Over the past months, we at The Fix have been covering various angles of generative AI – how to report on it, how to use it, and how it can impact news media. Check the “From The Fix” section for a collection of guides that will help you better understand one of the defining stories of 2023.
