Mathias Doepfner, CEO of German publishing house Axel Springer, spoke at the recent NOAH tech conference, where he said that the advent of automated article generation and AI journalist programs will not take jobs away from real journalists anytime soon. According to Doepfner, the AI journalists being used right now are only able to handle simple, “data-heavy” articles that mainly consist of reporting on facts and figures, scores, or other dry material that mostly has little room for interpretation, background information, or other journalistic conventions. One example are simple articles reporting soccer scores and game highlights from the Lower German league, a task that Axel Springer actually uses AI for.
The job of a journalist is a complex and nuanced one, involving background knowledge, writing skill, a sensitivity to audiences, and in many cases – investigative capabilities. At the current level of the technology, these are things that AI programs are not quite capable of yet. Even when AI journalists begin to learn deeper skills, they will still lack the human touch that makes for relatable, easy-to-read reporting for quite some time, if they ever truly get it at all. Instead, AI journalists are currently being used to automate and scale tasks that are either monotonous and take up time that real journalists could use generating better content, or to take care of the kinds of simple, tedious tasks that journalists normally have to go through in pursuit of a good story, such as sorting through data to spot trends or looking for numerous sources reporting the same thing in order to fact check.
While good journalism is still beyond the scope of AI for now, AI programs are slowly becoming more creative and more humanlike. A good example of this is the concept of artificial general intelligence, in which AI programs are not built for any specific task, but are instead taught to learn, grow, and synthesize new knowledge and even their own hypotheses and analysis. Alphabet chairman Eric Schmidt said that AI taking over simple and low-level jobs is a logical next step for the technology, and all signs thus far are pointing to him being correct.