The latest tool created for detecting typed text by AI is called Binoculars. It was developed by researchers at the University of Maryland in the United States, and promises to reduce so-called “false positives.” These erroneous results relate to the detection of texts attributed to artificial intelligence, even though they were written by a human. If we rely on certain current detection tools, we discover that, for example Macbeth Shakespeare wrote with software!
This can make you smile. But students can be wrongly accused of cheating and receive an ineligible grade. Therefore it is necessary to have more reliable detection. Currently, the error rate was approximately 1% false positives. But using the scope tool, we're going to go down to 0.01%. But if it is reliable, It's still not guaranteed Therefore, it does not answer the students' question. Many universities have also stopped using this type of tool.
Accept uncertainty
In the end, we may never know Whether the text was written by artificial intelligence or not. OpenAI, the creator of ChatGPT, has begun work on the detection tool. He appears to be in a better position to do so. However, he ended up giving up. He also felt unable to come up with a tool that was 100% reliable.
So, we're not done with the debate over homework written using ChatGPT. Perhaps the education sector will eventually adapt. The school has done well in incorporating the calculator, favoring logic rather than raw score. Maybe you'll get there with ChatGPT too.
It is possible that a tool such as endoscopes will still be of interest to other sectors, where false positives are less important. For example, to sort what is published in business or on the Internet. This will help dispel some doubts, or simply take a step back from what we are reading.
