“Has this text been written by a human or an artificial intelligence?” You have probably asked yourself this question on some occasions when seeing a post on social media or reading an article. The use of AI like ChatGPT has become normalized in a very short time, and many people are now using it daily to complete school assignments or even cheat on exams.
“When it comes to determining whether a text has ‘humanity’ or not, one of the most commonly used detectors is OpenAI’s text classifier, the company behind ChatGPT. However, the company seems to be dissatisfied with the results of their tool and has decided to shut it down until further notice.
In an update to a blog post, OpenAI announced the closure of its text classifier due to its low accuracy rate and stated that the services of this tool ended on July 20th.”
“We are working on incorporating feedback and currently researching more effective provenance techniques for text, and we are committed to developing and deploying mechanisms that allow users to understand whether the content is generated by AI,” explains OpenAI in its update.
The text classifier has given OpenAI more than one headache due to the high number of false positives it produced. These flaws are very similar to those exhibited by ChatGPT itself, which is capable of fabricating information and even lying on many occasions.”
Recently, OpenAI has been denounced after ChatGPT falsely claimed that a famous radio host in Georgia (USA) had been charged with fraud and embezzlement. In addition, a U.S. lawyer who used ChatGPT for a case ended up being sanctioned after presenting court cases invented by the AI at trial.
Some of the links added in the article are part of affiliate campaigns and may represent benefits for Softonic.
