ChatGPT detector tool made by 22-year-old in Toronto

OpenAI’s ChatGPT, an artificial intelligence chatbot tool, was a hot topic when the New York City Department of Education decided to block access to it on its networks and devices, citing concerns about the potential “negative impacts on student learning” and “concerns regarding the safety and accuracy of content.”

Now, a computer science student has developed a tool that could eliminate the need for a ban of the chatbot in school settings.

Toronto’s Edward Tian is a 22-year-old computer science major at Princeton University who knows about how people can misuse the AI-powered chatbot and decided to create a solution. Tian has spent “the last couple years” studying GPT-3 and other artificial intelligence tools that produce human-like text.

Upon becoming viral in November last year, educators globally feared that students would submit essays generated by the chatbot and there would be no way for teachers to detect plagiarism. To combat that, Tian developed GPTZero in Toronto, an app that can tell if a piece of text was written by AI or a human.

“Everyone deserves to know the truth and everyone deserves a tool at their fingertips that can determine whether something is human or machine-generated,” he told CTV News.

The way GPTZero works is that it measures the “perplexity, creativity, and variability” of a piece of text, and subsequently displays a score that reveals whether the text was generated by ChatGPT or a human. The application went live on January 3rd, and since, more than 300,000 people have tried it. “It was totally crazy. I was expecting a few dozen people,” Tian told CTV News.

If you’re looking to distinguish between a human-written and an AI-written piece of text, check out Tian’s app here.

Image credit: GPTZero

Via: CTV News