Former Special Assistant to President GW Bush Pippa Malmgren Responds to Artificial Intelligence Technology and Russia-Ukraine Conflict on ‘Making Money’ Theme
OpenAI, the San Francisco-based research firm behind ChatGPT, says it has released a new tool to help distinguish between AI text and human-written text.
The company announced on Tuesday the release of an initial version of the publication and said it aims to collect feedback and share improved methods in the future.
The creators of ChatGPT warned that it is impossible to reliably recognize all texts written by AI. However, the company believes that good classifiers can help flag automated misinformation campaigns, position an AI chatbot as a human, and leverage AI tools for academic dishonesty.
The launch of the Al Text Classifier comes after weeks of discussions in schools and colleges over fears that ChatGPT’s ability to write just about anything on command could fuel academic dishonesty and hinder learning.
POTENTIAL GOOGLE KILLERS COULD CHANGE OUR WORKFORCE AS WE KNOW IT
Teens and college students were among the millions of people who started experimenting with ChatGPT after it was launched as a free application on OpenAI’s website on November 30th. And while many found ways to use it creatively and harmlessly, the ease with which it could answer test questions and help with other tasks caused some educators to panic.
As schools opened for the New Year, New York City, Los Angeles, and other major public school districts began blocking its use in classrooms and on school equipment.
A logo of ChatGPT as seen on a mobile phone. (CFOTO/Future Publishing via Getty Images/Getty Images)
The longer a text passage is, the better the tool can recognize whether an AI or a human has written something. Input any text – a college admissions essay or a literary analysis of Ralph Ellison’s “Invisible Man” – and the tool flags it as either “very unlikely, unlikely, unclear if, possibly, or probably” is AI-generated .
CLICK HERE TO GET THE FOX BUSINESS APP
But much like ChatGPT itself, which has been trained on a vast trove of digitized books, newspapers, and online writings but often self-consciously spews untruths or nonsense, how it came to a conclusion is not easy to interpret.
The Associated Press contributed to this report.