Can ChatGPT Really Resolve Plagiarism Issue? Know How

Despite being a fantastic tool, ChatGPT has a negative aspect: plagiarism. This cutting-edge AI service can write with authority on practically any subject. Testing and grading are rendered invalid when students who are supposed to be proving their knowledge and comprehension of a subject use ChatGPT to cheat. Although having AI skills is great, students need also to study other things. The challenge of policing this issue has emerged. It's practically hard for a teacher to tell whether an essay was written by a student or a machine because ChatGPT was trained on such a large dataset of human writing. In an effort to identify text produced by AI, a number of tools have been developed, but their accuracy was too poor to be of any value. Photo Credit: MintLounge Business Insider notes that OpenAI is attempting to address this issue despite growing worries from instructors and restrictions on kids using ChatGPT. Earlier this week, Tom Goldstein, an associate professor of machine learning at the University of Maryland, tweeted about how effective ChatGPT's watermarking would be at being detected.

#OpenAI is planning to stop #ChatGPT users from making social media bots and cheating on homework by "watermarking" outputs. How well could this really work? Here's just 23 words from a 1.3B parameter watermarked LLM. We detected it with 99.999999999994% confidence. Here's how ? pic.twitter.com/pVC9M3qPyQ

— Tom Goldstein (@tomgoldsteincs) January 25, 2023

This debate could be easily resolved, and any worries could be allayed, if a program could detect plagiarism with almost 100% accuracy. Making the large language model (LLM) select terms from a constrained vocabulary in order to create a blacklist of phrases that the AI is not permitted to use and a whitelist of words that are permitted, according to Goldstein, is one approach. A sample would be considered artificial intelligence-generated if an excessively high number of words from the whitelist appeared in it. Since working one word at a time, as most LLMs do, makes it difficult to anticipate which words could be necessary for a debate, this very simplistic technique would be too constrictive. In order for ChatGPT to plan a phrase that can contain just terms from the white list while still making sense, Goldstein advises giving it the capacity to look ahead more than one word. Photo Credit: CN The introduction of ChatGPT into the community writing pool caused quite a stir, and it has the potential to be an excellent instructional tool as well. Artificial intelligence should be taught in schools because it will undoubtedly be a crucial technology to comprehend in the future. However, until the issue of plagiarism is resolved, there will be debate regarding its proper teaching in schools.