UNITED STATES: OpenAI, the developers of a ChatGPT bot that caused a sensation for its ability to replicate human writing launched a new program on Tuesday. The newly introduced tool intended to identify instances in which written works are produced by artificial intelligence.
Concerns about how the software might be used to help students with their assignments and cheat on exams sparked a lot of discussion at schools and universities in the United States and other countries at the time of the announcement.
OpenAI’s ChatGPT adds detection tool
Its detection tool has been trained “to identify between text produced by a person and text generated by AIs from a range of suppliers,” according to a blog post posted by the US business OpenAI on Tuesday.
The bot from OpenAI, which just received a large cash inflow from Microsoft, answers straightforward requests with reams of text motivated by information acquired online.
OpenAI issued a warning that its tool has a chance of making mistakes, especially with messages that are less than 1,000 characters.
OpenAI wrote in the post that while it is hard to accurately identify all AI-authored content, “we believe effective classifiers might inform mitigations for erroneous claims that AI-generated language was written by a human.”
“For instance, operating automated disinformation campaigns, utilising AI tools for academic dishonesty, and passing off an AI chatbot as a human,” the statement read.
The first such prohibition at a college in the nation was implemented last week by a prestigious French university, which barred students from using ChatGPT to do coursework.
The choice was made soon after it was reported that ChatGPT had succeeded in tests at a US law school after composing essays on subjects as diverse as constitutional law and taxation.
Although ChatGPT continues to produce factual errors, educational institutions have hurried to outlaw the AI tool.
OpenAI wrote in the post, “We know that detecting AI-written content has been a significant topic of discussion among educators, and as important is recognising the limits and consequences of AI-generated text classifiers in the classroom.”
We are interacting with educators in the US to find out what they are witnessing in the classrooms and to talk about ChatGPT’s strengths and weaknesses. Its usage in educational institutions has been outlawed by authorities in New York and other places.
According to a consortium of Australian universities, they will alter exam forms to outlaw AI technologies because they are considered cheating.
The classifier performs worse in other languages, according to OpenAI. Hence it is advised to only use it with English content.
Also Read: Earth to Exceed 1.5°C Warming Limit in the Next Decade, Claims New AI Study