A software tool for identifying text generated by artificial intelligence has been released by OpenAI, the company behind the well-known chatbot ChatGPT, the company announced in a blog post on Wednesday.
ChatGPT is a free program that responds to a prompt by writing articles, essays, jokes, and even poetry. Since its launch in November, it has gained a lot of popularity, but it has also raised concerns about copyright and plagiarism.
The AI classifier is a language model that was trained on a dataset of text written by humans and AI on the same subject. Its goal is to tell the difference between text written by humans and text written by AI. According to the company, it addresses issues like academic dishonesty and automated misinformation campaigns by utilizing a variety of providers.
OpenAI acknowledges in its public beta mode that the detection tool is extremely unreliable for texts with less than 1,000 characters and that AI-written text can be edited to fool the classifier.
OpenAI stated, “We’re making this classifier publicly available to receive feedback on whether imperfect tools like this one are useful.”
“We recognize that identifying AI-written text has been an important topic of discussion among educators, and we also recognize the limitations and effects of AI-generated text classifiers in the classroom,” the authors write.
Some of the largest school districts in the United States, including New York City, have banned the artificial intelligence chatbot since its November debut due to concerns that students will use the text generator to cheat or plagiarize.
GPTZeroX is one example of a third-party detection tool developed by others to assist educators in identifying AI-generated text.
OpenAI stated that it will continue to work on the detection of AI-generated text and is meeting with educators to discuss ChatGPT’s capabilities and limitations.
