After the New York Department of Education decided to block the use of the ChatGPT artificial intelligence chatbot in the city’s teaching environment, OpenAI decided to take “mitigating measures.” to help anyone determine if a text was generated by software. Among these measures may even be the technique of watermarking. TechCrunch.
In response to pressure from North American teachers who discovered that students were using the tool to do their homework, and who expressed concerns about how these language models might encourage plagiarism and cheating in school assignments, OpenAI, the initiative behind ChatGPT, assured: “We always want transparency regarding the use of AI generated text”.
Since this “intimacy” isn’t something everyone can expect from all users, they released a memo in which they promised to “work with educators on helpful solutions and other ways to help teachers and students benefit from AI.” a means of separating human-produced writing from machine-produced.
What worries most teachers about using ChatGPT?
In an article published on the website living roomGary Smith, professor of Economics at Pomona College, warns of the “misleading” character of the GPT-3’s true intelligence, Generative Pre-Trained Transformer 3, the technical name for the current version of ChatGPT.
“What it does is relatively simple: the GPT-3 manages to deliver remarkably intelligible speech and write persuasive essays, stories and even research papers,” says the professor. Despite everything, “He has no idea what the words mean” assures Smith.
But the biggest problem with ChatGPT is cheating and this can be proven as follows: Numerous AI offers for writing school assignments are currently available on Google. To prove this, recently published reports The Wall Street Gazette and on forbes completed three university papers in minutes.
Source: Tec Mundo
