Some of my teachers from the Faculty of Law, from the National University
Autónoma de México (UNAM), are concerned about being able to detect if their
students are already using ChatGPT for the use of essays, book reviews,
tasks in general, recently a Princeton University student has
created an application that helps to detect if a text has been written by a human being
or through the artificial intelligence tool ChatGPT.
Edward Tian, a 22-year-old computer science and journalism student at the University of
Princeton, is the creator of GPTZero, a free and publicly accessible tool capable of
to detect if a text has been written by a human being or by a machine.
The algorithm behind his app, called GPTZero, can “quickly and quickly detect
effectively if an essay, or article, is written with ChatGPT or by a human”. The
beta version of the application is available for free.
Tian, an excellent student at this North American university, is concerned about the
ethical dilemmas that chatbots pose, as well as what he describes as nature
of "black box" of the great linguistic models such as ChatGPT.
The opaque nature of the models causes people to misunderstand them and therefore
use incorrectly. Generating vices in the educational formation of the
students.
The apparent brilliance of ChatGPT distracts attention from major flaws: it can
produce harmful content and often write fiction as if it were fact. Besides,
that fails to verify the sources of information according to reports from
own media that have decided to use it, as a test, in tasks
of edition.
Due to these limitations and the possibility of cheating, many teachers are
concerned about the impact of ChatGPT in the classroom. The app has already been banned
by school districts across the northern country, such as those in New York, Seattle, and Los
Angels.
Since the use and popularity of ChatGPT has increased recently due to its
ability to generate coherent essays on virtually any topic in
matter of seconds, this student Edward Tian, was given the task of creating this
detection tool. For its part, this technology has aroused the interest of
investors, and the Wall Street Journal reports that parent company OpenAI could
soon attract investments that value it at 29,000 million dollars.
As part of their AI detection scheme, they have released a new GPT detector
writing and AI chat. The new software is said to be 97% accurate and a
false positive rate of only 1%. Unfortunately, the detector is not available yet,
but it is expected to go on sale in April this year.
So the only way to put limits on this Pandora’s box of technology is to
wait for the growth of ChatGPT itself, and prohibiting its use and disseminating this
types of detection tools, at least in terms of their more
harmful.