In my October column for FTChinese, I explored the outsourcing of intellectual labor to GenAI. While these tools generate text effortlessly, they often produce what researchers call AI workslop, i.e., superficially fluent but substantively nonsensical content. Such productivity may help writers overcome writing bottlenecks, but it may threaten media outlets’ core competencies and business models.
As a writer myself, my two actionable recommendations:
1. Grasping the "first" and the "last." The hardest first draft remains valuable because it forces journalists to determine what to say and how to pitch the story. Even though AI can inform story ideas, the final decision is a human’s own discretion. I am intrigued by several recent conversations with industry friends—they expressed strikingly similar ideas: "ping ponging ideas with GenAI" 🏓 (other similar words include: using AI "interactively," "iteratively," "back and forth," and "battling around 300 rounds", if you know traditional Chinese martial arts and kung fu movies.
So when do you stop THAT iteration? That moment is your unique intellectual signature, reflecting your taste, your restrictions, or the long-term norms you have cultivated from your family and community. I see these as assets to be leveraged.
2. Training one's intellectual taste and "creativity judgment" has become unprecedentedly critical. In my "GenAI for Multimedia Authoring" - a new undergraduate course at the
College of Liberal Arts and Social Sciences, CityUHK- I adapt a classic exercise (common in any journalism and media curriculum, as practiced in comparing different news outlets covering the same international disputes): asking students to compare and pick the best version from multiple GenAI-generated news titles. In one exercise, students use Gemini to create six versions of RedNote news titles about the recent US-China tech rivalry, and students picked one and said, "That's like the vloggers in the Chinese mainland". But, how so?
I argue that GenAI is not helping or freeing us just "in the last mile." It engages us even before we write down the first words, as well as in every step of decision-making. That is something that warrants more observations on human-AI collaborative creativity.