Ask the new AI program ChatGPT to write an essay on the causes of the American Civil War and watch it generate a compelling piece in seconds.
It’s one reason New York school officials have begun banning this influential but controversial writing tool capable of producing human-like paragraphs of text.
The largest school district in the United States has decided to ban all types of devices and the use of the ChatGPT website on its school network, which could affect other schools and teachers, to prevent students from cheating. are struggling. The makers of ChatGPT say they are also looking into ways to detect inappropriate use.
This free tool has only been on the market for a few weeks, but it’s already raising tough questions about the future of AI in academia, the tech industry and many businesses.
What is ChatGPT?
The ChatGPT program launched on November 30, but is part of a broader set of technologies being developed by OpenAI, a San Francisco-based startup that has close ties to Microsoft.
It is part of a new generation of artificial intelligence (AI) systems that can hold conversations, generate readable text on demand, and even novel images based on what they “learned” from a vast database of digital books. And can also produce video, written online and other media.
But unlike previous versions of so-called “big language models” like OpenAI’s GPT-3, released in 2020, the ChatGPT tool is freely available to anyone with an Internet connection and designed to be easy to use. It acts like a written dialogue between the AI system and the person asking it questions.
Millions of people have played with ChatGPT over the past few weeks, using the program to compose or write silly songs or poems, to help you spot mistakes, or for more practical purposes, such as composing emails. for. All those inquiries are also helping you to become smart.
What are its disadvantages?
Like other similar systems, ChatGPT can produce believable prose, but that doesn’t mean that what it says is true or logical.
Its release came with little guidance on how to use it, other than a promise that ChatGPT would admit when it was wrong, challenge “false premises” and reject requests intended to generate an offensive response. Since then, however, its popularity has prompted its creators to try to temper some people’s expectations.
“ChatGPT is incredibly limited, but good enough at some things to create an illusory impression of greatness,” Sam Altman, CEO of OpenAI, tweeted in December.
Altman acknowledged, “It’s a mistake to trust him for something important right now.”
“This is a preview of progress. We have a lot of work to do in terms of robustness and veracity.”
Can CHATGPT be used to write school essays?
ChatGPT himself, when asked by The Associated Press, asked the question to be answered in capital letters to make it appear as if it was a school principal asking him to issue a short, sharp warning over a public address system. went. school:
“Do not use ChatGPT or any other automated writing program for school essays. This is cheating and will not be tolerated. If they are caught using ChatGPT or any other automated writing program for school essays then its serious There will be consequences.”
But when asked to answer the same question, ChatGPT offered this more measured warning: “As a general rule, it is not advisable to use ChatGPT or any other writing tool for school essays, because This is considered cheating and it is not going to benefit the students in the long run.”
What do schools say?
Many school districts are still having trouble setting policies on whether and how the program can be used.
The New York City Department of Education said on January 5 that it was restricting access to the program through school networks or devices, saying it feared negative effects on student learning, as well as “security issues”. And concern about the accuracy” content”.
However, nothing stops a student from accessing ChatGPT from a personal phone or home computer.
School spokeswoman Jenna Lyle said, “While the program may provide quick and easy answers to questions, it does not encourage critical thinking and problem solving skills, which are essential for success in school and life.”
Written by a Human or by Artificial Intelligence?
“To try to determine whether something was written by a human or by an AI, you have to be attentive to the absence of personal experiences or feelings, look for inconsistencies in writing style and look for filler words or repeating phrases.” Observe the usage. . These may be indications that the text was generated by AI.”
That’s what ChatGPT told an Associated Press reporter when asked how to tell the difference.
The OpenAI firm said in a statement in early January that it plans to work with educators while also learning how people experience ChatGPT in the real world.
The company said, “We don’t want ChatGPT to be used for deceptive purposes in schools or anywhere else, so we are already developing mitigation measures to help anyone identify ChatGPT-generated text.” “
Does it threaten Google?
There has been speculation that ChatGPT could replace Google’s currently dominated Internet search business, but the tech giant has been working on similar technology for years, but Google has been more cautious about the idea of launching it. public.
In fact, it was Google itself that helped spur the trend of ever bigger and smarter AI language models that can be “pre-trained” for large amounts of writing. In 2018, the company introduced a system called BERT, which uses a “transformer” technique by which it compares words in a sentence to predict meaning and context. Some of those advances are now integrated into Google Search.
However, there is no doubt that successive versions of these programs have had an impact. Microsoft has invested at least $1 billion in OpenAI and has an exclusive license to use GPT-3.