ChatGPT-a conversational language model which was launched in November and is free and simple to use - can swiftly produce poems, math equations or essays on topics, bringing concern that students will misuse the technology. And because it doesn't copy an existing text, there is no easy way to be certain whether a human or a bot wrote the answer.
As many educators began to worry about whether students may use ChatGPT to generate papers, Edward Tian had an idea. The 22-year-old student, who studies computer science and journalism, decided to build an app to detect whether a text- was human-written or AI-written.
Over a few days in a Toronto coffee shop during winter break, he got to work. On Jan. 2, he launched' GPTZero. It analyzes' different properties of a text for its "perplexity", which is the randomness of the text, and the "burstiness", 'which is the variation of the text over time. So, a human-written text would have high perplexity, something very unfamiliar to an Al model, and exhibit properties of burstiness, which are non-common items that appear in random clusters （集 群）, rather than being uniformly distributed.
Tian said he expected a few dozen people to ever try it. But he woke up the next morning stunned by the response. He has even heard from people all over the world— many of them are teachers or college admissions officers, Many people have subscribed for updates from Tian as he works to improve the technology.
The quick response to Tian's effort highlighted the breakneck pace at which technology is changing classrooms, teaching, and the ways that people define and understand learning. Tian believed everyone deserved to reap the benefits of Al, but safeguards were needed to make sure new technologies were not abused.
Tian said it would be sad if, years from now, people mostly relied on AI and writing became far more uniform. "There's something implicitly（含蓄地） beautiful in human prose, " he said, "that computers can never copy. "
