Edited By: Bharat Upadhyay
Last Updated: February 01, 2023, 12:45 IST
The AI Text Classifier is a fine-tuned GPT mannequin
The AI Text Classifier is a fine-tuned GPT mannequin that predicts whether or not a chunk of textual content was generated by AI from quite a lot of sources, resembling ChatGPT.
After launching the preferred AI chatbot — ChatGPT, the Microsoft-backed AI analysis and deployment firm OpenAI has now launched a brand new software referred to as ‘The AI Text Classifier; which can detect AI-generated content.
The AI Text Classifier is a fine-tuned GPT model that predicts whether a piece of text was generated by AI from a variety of sources, such as ChatGPT. “This classifier is available as a tool to spark discussions on AI literacy,” the company said on the new tool’s web page.
According to the corporate, AI Text Classifier’s reliability sometimes improves because the size of the enter textual content will increase. Compared to a beforehand launched classifier, this new software is considerably extra dependable on textual content from newer AI techniques.
Also, the classifier has plenty of vital limitations. It shouldn’t be used as a main decision-making software, however as an alternative as a complement to different strategies of figuring out the supply of a chunk of textual content. The classifier could be very unreliable on quick texts (beneath 1,000 characters). Even longer texts are generally incorrectly labeled by the classifier.
Sometimes human-written textual content will likely be incorrectly however confidently labeled as AI-written by our classifier. “We suggest utilizing the classifier just for English textual content. It performs considerably worse in different languages and it’s unreliable on code,” the website reads.
Text that is very predictable cannot be reliably identified. For example, it is impossible to predict whether a list of the first 1,000 prime numbers was written by AI or humans, because the correct answer is always the same. AI-written text can be edited to evade the classifier.
“Classifiers like ours can be updated and retrained based on successful attacks, but it is unclear whether detection has an advantage in the long-term. Classifiers based on neural networks are known to be poorly calibrated outside of their training data. For inputs that are very different from text in our training set, the classifier is sometimes extremely confident in a wrong prediction,” the corporate mentioned.
Read all of the Latest Tech News right here
Source web site: www.news18.com