After a successful run of their AI-generated content, OpenAI now launches its AI-generated content detector.
ChatGPT, the chatbot that uses artificial intelligence to generate content, has taken the internet by storm. The chatbot is being used by a majority of the workforce globally, regardless of fields and purposes. This vast use of ChatGPT has also figured out issues related to it, such as plagiarism, harmful and misleading information, limited contextual awareness, and most definitely a lack of acknowledging a prompt or command. To detect the loopholes in its generated content, OpenAI itself launched the AI Generated Content Detector, which will be known as the “GPT-Classifier.”
How GPT-Classifier will detect AI Generated Content?
According to OpenAI, the GPT-Classifier comes as an extension to their GPT to get better results. The “GPT-Classifier” will be able to detect if a text was written by a human or by using artificial intelligence. The GPTs now available on the web are mostly trained to take input of any sort of text, while the GPT-Classifier of OpenAI is “fine-tuned on a dataset of pairs of human-written text and AI-written text on the same topic.”
The classifier will judge the input and provide a final output by comparing a piece of text written by AI and humans. If the input matches any AI content, it will show the details of the detection result, which include the source, author, and GPT used.
Can I use GPT-Classifier now?
The availability of “GPT-Classifier” might sound like music to the ears of teachers, research instructors, team leads at the corporate houses, and obviously to the content editors. The use of GPT-Classifier is very limited now since it can primarily detect only 26% of AI-generated text while showing 9% of human-written content as AI-generated content. So there are still miles to go before the GPT-Classifier can make a real impact on this scenario.
How to use ChatGPT?
ChatGPT can be used by browsing OpenAI’s website. You can directly go to this to sign up for ChatGPT. You may face an error to enter the site as millions of people jams their server, in such cases keep refreshing until the login and Sign Up prompt appears. Once you are signed up or logged in, click +New Chat to start chatting and commanding.
Kindly be cautious not to trust ChatGPT-generated content without fact-checking. Since it is Artificial Intelligence, it appears to be showing misleading info in a number of cases.
