Tech website CNET tried to use ChatGPT to write articles, but the quality and accuracy left much to be desired.
ChatGPT is the latest conversational AI from OpenAI. The software has gained a massive following thanks to it being one of the most advanced conversational AIs yet released. ChatGPT has been used in a variety of applications and scenarios, but CNET is one of the biggest websites to try using the AI to write articles…an experiment that did not go well.
As a result of numerous issues, CNET had to publish multiple corrections to articles originally written by the AI. In a statement on their website, CNET editor Connie Guglielmo explained the outlet’s use of ChatGPT was in line with the company’s commitment to testing new technology.
The Washington Post had a slightly different take, calling CNET’s experiment “a journalistic disaster.” The Post also pointed out that CNET only admitted to using ChatGPT after it was called out by sharp-eyed users and other sites.
On Tuesday, CNET began appending lengthy correction notices to some of its AI-generated articles after Futurism, another tech site, called out the stories for containing some “very dumb errors.”
As the Post points out, the use of AI for journalism also brings up plagiarism issues, since many AIs remix other articles and sources, without properly attributing them.
If CNET’s experience is any indication, AI still has a way to go before it can replace quality journalism and writing.