So far, newsrooms have pursued two very different approaches to integrating the latest AI tool, ChatGPT, into their work. Technology news site CNET has secretly started using ChatGPT to write entire articles, only for explosive testing. It was eventually forced to issue a correction amid allegations of plagiarism. Buzzfeed, on the other hand, has taken a more measured, cautious approach. Its leaders want to use ChatGPT to generate quiz answersGuided by journalists creating topics and questions.
You can boil these stories down to a fundamental question many industries are currently facing: How much of an AI system should we control? CNET gave too much and ended up in an embarrassing mess, while Buzzfeed’s more cautious (and transparent) approach to using ChatGPT as a productivity tool was generally well received. and caused its stock price to skyrocket.
But here’s the dirty press secret: Charlie Beckett, a professor at the London School of Economics who runs a program called JournalismAI, says a surprisingly large amount of that can be automated. Journalists frequently reuse text from news outlets and steal ideas for stories and sources from competitors. It makes perfect sense for newsrooms to explore how new technologies can help them make these processes more efficient.
“The idea that journalism is a blooming flowerbed of originality and creativity is absolutely rubbish,” says Beckett. (Oi!)
It’s not necessarily a bad thing if we could outsource some boring and repetitive parts of journalism to AI. In fact, it could free up journalists to do more creative and important work.
A good thing For example I’ve seen this using ChatGPT to repackage text on the news wire into the “smart succinct” format used by Axios. The chatbot seems to have done its job well enough, and I can imagine that any journalist responsible for imposing that format would be happy to have time to do something more interesting.
That’s just one example of how newsrooms can successfully use AI. AI can also help journalists summarize long passages of text, skim through datasets, or come up with ideas for headlines. In the process of writing this newsletter, I myself used a number of AI tools, such as word processing autocomplete and audio interviews.
But there are some major concerns about the use of AI in newsrooms. An important issue is privacy, especially around sensitive stories where it is important to protect your source identity. This is a problem that journalists at MIT Technology Review have encountered when using audio transcription services, and sadly the only way to solve this problem is to transcribe sensitive interviews according to their needs. manual way.
