An algorithm designed by OpenAI can identify if ChatGPT is being utilized for essay or research paper writing.
When it comes to identifying AI-written text, ChatGPT’s technology has 99.9% accuracy.
OpenAI’s New Algorithm Can Identify Use of ChatGPT For Essay or Research Paper
The technology has been available for a year, but internal discussions and worries about user feedback have prevented it from being launched.
An OpenAI survey found that nearly one-third of devoted ChatGPT users disapproved of the technology.
There are worries that non-native English speakers could be unfairly affected by the instrument.
Despite the tool’s technical promise, OpenAI is considering alternatives and analyzing the risks.
The watermarking technique used by the anti-cheating technology modifies the token selection in text production in a subtle way.
The watermarking produces an undetectable pattern that OpenAI’s system can identify.
There are worries that methods like translation or emoji addition and subtraction could be used to erase the watermark.
It’s difficult to decide who should get access to the detector.
59% Middle and High School Instructors Reveal Use of AI in Homework
Profs and educators are excited about resources to counteract the misuse of AI in the classroom.
59% of middle- and high-school instructors surveyed said they thought students have utilized artificial intelligence (AI) for homework.
Concerned about how the anti-cheating tool may affect fairness and user experience, OpenAI is being cautious.
Watermarking solutions are also being developed by other companies, like Google, which is now testing its SynthID tool.
Since audio and visual watermarking have a greater impact than text, especially in election years, OpenAI gives them priority over text.
In order to get public opinion on the tool, OpenAI has conducted polls and engaged in internal debates.
False allegations and their possible effect on ChatGPT’s performance are causes for concern.
In an effort to conform to public sentiment and prospective new legislation regarding AI transparency, OpenAI keeps looking into less contentious options.
