The innovative ChatGPT chatbot, created by OpenAI, has become extremely popular, but this has caused a few problems. Many people, including writers, coders, and marketers, are using this AI-powered tool to produce content, resulting in a wave of AI-created material that companies are facing.
Plagiarism and unoriginal material are frequent topics of conversation when discussing ChatGPT and its products. Therefore, OpenAI has created GPT-Classifier, a tool that can identify and inform whether a piece of text has been made using ChatGPT or another GPT tool.
OpenAI Introduces GPT-Classifier for Evaluating Text Composed by Artificial Intelligence
OpenAI has made a statement revealing their new AI classifier that can detect AI-created text as the focus on AI-writing and AI-generated text increases and more businesses are utilizing ChatGPT output.
The latest GPT-Classifier seeks to decide if a given piece of text is generated by an AI or written by a human. While ChatGPT and related GPT models have been trained on a variety of text types, the GPT-Classifier has been “tailored to a group of pairs comprising of human-created text and AI-crafted text on the same topic.”
Rather than gathering data and reproducing its most accurate version, the GPT-Classifier makes an effort to compare the similarities between texts that have been written by humans and those made by AI to recognize any disparities that disclose the author.
Currently, the abilities of GPT-Classifier to identify things are restricted.
Editors and educators will be delighted to hear that it is possible to easily identify AI-generated text, however, OpenAI has warned that their “classifier is not completely dependable.”
The GPT-Classifier has an accuracy rate of 26% when correctly recognizing AI-generated text. Additionally, OpenAI has observed that the tool’s accuracy improves with the length of the input text, with 9% of human text being labelled as AI-written.
A brief examination of the GPT-Classifier revealed a sample of human origin and the AI classified it as “very unlikely to be AI-generated,” and accurately identified a ChatGPT-generated phrase about USB problems as “possibly AI-generated.”
There is an Elevated Requirement for ChatGPT Detection
Despite the GPT-Classifier’s restrictions, there is likely to be a great demand for dependable ChatGPT recognition. OpenAI’s dedication to constructing and releasing a free GPT detection tool is essential because as more individuals, such as students, authors, programmers, and others, utilize AI-text generation tools, being aware of and detecting this input will be necessary.