Artificial intelligence won’t replace you, but someone using AI just might.
This was a pressing topic for attendees at “The New and Old Newsroom Skills Students Need in the Era of Generative AI” panel, which covered the ethics of utilizing AI for writing as well as the many pitfalls AI must overcome to craft nuanced journalism.
The panel featured Laura Davis, founder of the AI writing tool Stylebot, Rebecca Haggerty, undergraduate journalism director at the University of Southern California, Laura Castañeda, professor of professional practice at USC and Tomoki Chien, a USC student.
Generative AI can create ideas, summarize facts and produce writing at the click of a button. However, it can also falsify quotes, invent untrue facts and repeat the biases of data it was trained on.
Castañeda asked the room of attendees if they believed AI can be racist, ageist or ableist. She cited the AI blunder of an Asian American MIT student asking Playground AI to make her LinkedIn photo appear professional–only for her skin to be lightened and her eyes to be turned bright blue.
But AI can also be a useful tool in the newsroom for synthesizing information, generating headlines and seeing patterns in large data sets.
“It’s an interesting way to think about how we interact with information online, it’s not the way we have interacted with info in the past,” Haggerty said. “Making things up and being false is a hallmark of generative AI tools.”
The panel emphasized that human oversight is required when publishing stories. In the age of AI, having a sharp eye for copy editing is more critical than ever.
While AI can write a hard news story, it struggles with bringing an emotional core to a narrative. Technology lacks the creativity, innovation and judgment needed for quality journalism.
“I have not seen flashes of brilliant or particularly good writing,” Haggerty said.
However, quality writing did not stop Gizmodo en Español editors from being fired and replaced with AI translators in August.
Chien highlighted the popularity of classmates utilizing AI for general education college courses. ChatGPT easily summarized five points on the letters of Paul for a midterm on the New Testament.
“Like Google, it’s a tool and we have to learn how to use it. I would prefer that my students learn how to write and don’t just hand it over to an AI generator,” Castañeda said, sharing an excerpt of her syllabus detailing policies for AI-generated work. “They can use AI as an outline for brainstorming and to check AP style and AP citations. However, they have to fact-check everything that they submit to me.”
Attendees broke into small groups to discuss their comfort levels with students using generative AI. Students and professors paired to share their opinions on the blurred distinction between cheating and embracing technology.
For now, students and journalists must learn to adapt to the tool. AI is embedded in our culture, and it’s here to stay.