chatgpt purple color design

3 ways journalists can use AI writing ethically

This was originally a video, post and thread elsewhere.

ChatGPT is especially scary to journalists because AI writing feels a lot like cheating. But it doesn’t have to be. Here are three ways I’ve experimented with AI writing that don’t break any journalism codes:

Shorten: Drop your existing draft news article into ChatGPT and request it to shorten it using as much of your existing language as possible. This is using your words and your work, basically a copy-edit that helps save your own editor.

Draft: Drop in an interview transcript and request a summary with key points, or even a draft article. You’ll adapt it and integrate it with other interviews, and it uses language from your interview, so it’s still original work. I have also trialed translating my stories into other languages, and we’ve been told the accuracy was relatively good.

Summarize: Many press releases and reports end up primarily a nut graf. We’re looking for key facts that defend the people stories we’re including. Request a paragraph, fact check and rework it.

The act of writing is often a process of understanding. I am not telling you not to do that! This is a tool!

AI writing isn’t a super intelligent being (yet), it’s access to unlimited interns

Use AI to focus on the people stories, beautiful language or making connections across different industries or places. Remember: our words are a tool to make connections with people. I welcome anything that helps us with that! Good luck!

Leave a Reply