ChatGPT’s owner has released an “imperfect” AI-generated text detection tool

TORONTO, Canada (Reuters) – OpenAI, creators of the popular chatbot ChatGPT, has released a tool for selecting AI-generated text, the company said Wednesday.

ChatGPT is a free program that generates text in response to a request, including essays, essays, jokes, and even poetry, which has gained huge popularity since its debut in November, with copyright and plagiarism concerns raised.

The AI ​​classifier, which is a language model trained on a data set of text pairs, aims to distinguish text written by bots. The company said the software uses a variety of sources to address issues such as automated disinformation campaigns and academic hoaxes.

In public beta mode, OpenAI realizes that the detection tool is not very reliable on text less than a thousand characters and that the text typed by the AI ​​can be edited to fool the classifier.

“We’re making this workbook publicly available to get feedback on whether imperfect tools like this one might be useful,” OpenAI said.

“We recognize that the identification of written text by AI has been an important point of discussion among educators, and equally important is the recognition of the limitations and impacts of software-generated text classifiers in the classroom.”

Since ChatGPT debuted in November and gained widespread popularity among millions of users, some of the largest school districts in the US, including New York City, have banned the chatbot over concerns that students might use the software to cheat or plagiarize.

OpenAI said it is engaging with educators to discuss the features and limitations of ChatGPT and will continue to work on AI-generated text detection.

signed on 247And pix supportAnd Subscribe to TV 247in the channel 247 – Nada And watch:

Knowledge is liberating. I want to be a member. Follow us cable.