Odaily Planet Daily News OpenAI has added a watermark system to the text created by ChatGPT and prepared a tool to detect watermarks, which has been prepared for about a year. But there are disagreements within the company about whether to release the system because it may harm the company's profits. The company found that when there are enough AI texts, this method is "99.9% effective" in detecting AI texts, and in a survey it commissioned, "a quarter of respondents worldwide support AI detection tools." But OpenAI seems to be worried that the use of watermarks will make the ChatGPT users surveyed lose interest, and nearly 30% of them said that if watermarks are implemented, they will use the software less. Some employees have other concerns, such as using Google Translate to translate texts back and forth between different languages, or letting ChatGPT add emojis and then delete them, such tricks can easily prevent watermarks. Despite this, employees still believe that this method is effective. However, given the dissatisfaction of users, some of the methods suggested to try "may be less controversial among users, but unproven." (WSJ)