Online technology site CNET recently tried using AI ChatGPT to write some of its articles. Unfortunately, the result was a failure..
Readers were very disappointed by this attempt, especially from a site that is highly reputed in its field. This experiment is an interesting illustration of the fact that AIs are not yet ready to completely replace human work.
ChatGPT for editors: CNET’s initiative comes in for criticism
ChatGPT is an artificial intelligence designed to automatically generate text content from information provided by the user. The program analyzes the information and then produces a complete article on the requested topic. The AI answers are very pertinent, sometimes with illustrative examples.
While this kind of technology may seem promising, it doesn’t always work as expected. This is exactly what happened with CNET. The articles produced by IA were completely lacking in depth and contained glaring errors that disappointed readers. What’s more, they felt cheated because they thought the articles were written by qualified human journalists. Critics pointed out that the experiment resembled an attempt to eliminate the work of editors.
Did CNET want to make the most of its editorial budget?
Bottom line, CNET has acknowledged its error according to an announcement published by CNET editor-in-chief Connie Guglielmo. She said she had taken steps to verify ChatGPT-generated articles on the website. Fortunately for them, the scandal didn’t affect their image too much. Indeed, they acted swiftly to rectify the situation and showed thatthey take reader satisfaction seriously.
This failure underlines the difficulty of creating AIs capable of producing coherent and meaningful content without human intervention. While progress in AI is impressive, it’s clear that the task of replicating the creative work done by journalists remains very difficult to accomplish at the moment.