Shulou(Shulou.com)11/24 Report--

On January 17, with the rise of generative artificial intelligence technologies such as chat robot ChatGPT, many American universities began to adjust course contents, change teaching methods and other preventive measures.

Last month, Anthony Antony Aumann, a professor of philosophy at the University of Northern Michigan, read what he called "the best paper in the class" while correcting papers for a course he taught. This paper has clear paragraphs, appropriate examples and rigorous arguments.

Orman questioned his students whether he wrote the article himself. The student admitted to using ChatGPT. This kind of chat robot can use simple sentences generated automatically to convey information, explain concepts and generate opinions. In this case, it can be said that ChatGPT wrote this paper.

Auman was shocked by the discovery and decided to change the way he wrote his paper this semester. He plans to require students to write first drafts in class using browsers and computers with restricted access. Students must also explain each revision in a later draft. Aumanhai may also stop letting students write papers for the next few semesters and plans to integrate ChatGPT into the curriculum by evaluating chatbot answers.

"what happens in class will no longer be, 'here are some questions, let's talk about it,'" Orman said, but "things like, 'what is this robot thinking?'"

American university professors, department heads and administrators like Auman have begun to overhaul classroom teaching in order to deal with ChatGPT, which could lead to a huge change in teaching methods. Some professors are completely redesigning the courses they teach, introducing more oral exams, group discussion assignments and handwritten content assessments to replace papers.

These measures are a real-time response to the wave of new technologies brought about by generative artificial intelligence. Artificial intelligence laboratory OpenAI released ChatGPT in November last year, which is at the forefront of this technology wave. ChatGPT automatically generates logical text based on short prompts, and many people use it to write love letters, poems, fan novels, and even finish homework.

This has affected many middle and high school teaching in the United States, and teachers and administrators need to tell whether students are using chatbots to do their homework. To prevent cheating, some public schools in New York City and Seattle have banned the use of ChatGPT on campus networks and access devices, but it is easy for students to find workarounds to access ChatGPT.

But in the field of higher education in the US, colleges and universities have been reluctant to ban artificial intelligence tools because managers doubt whether it is effective and do not want to infringe on academic freedom. This means that teaching methods on American college campuses are changing.

"our overall policy should explicitly support the authority of teachers to manage courses," said Joe Glover, provost of the University of Florida, rather than specific ways of cheating. "it's not going to be the last innovation we have to deal with," he said. "

The starting point of Glover and others is that generative artificial intelligence is still in its early stages. OpenAI is expected to soon release another artificial intelligence tool, GPT-4, which is better at generating text than ChatGPT. Google has developed its own chat robot LaMDA, and Microsoft is discussing a further $10 billion investment in OpenAI. Silicon Valley startups such as Stability AI and Character are also postgraduate in artificial intelligence tools.

A spokesman for OpenAI said that the lab has realized that the developed program may be used to mislead the public and is developing technology to help people identify what content is automatically generated by ChatGPT.

In many universities, ChatGPT has now leapt to the top of the teaching agenda. Managers are setting up a working group on how to deal with ChatGPT and leading school-wide discussions, most of which are on how to adapt to generative artificial intelligence technology.

Professors at George Washington University in Washington, D.C., Rutgers University in New Brunswick, New Jersey, and Appalachian State University in Boone, North Carolina, are phasing out after-school assignments, a major academic curriculum assessment method that now seems vulnerable to chatbots. Instead, they began to choose more class assignments, handwritten papers, group assignments and oral exams.

Simple requirements such as "write five pages about this or that" have disappeared. Instead, some professors carefully designed questions that they thought were too smart for chatbots and asked students to write down their understanding of their lives or current events.

Sid Dobrin, dean of the English department at the University of Florida, says students "copy because homework can be plagiarized."

Frederick Louis Aldama, director of the humanities at the University of Texas at Austin, revealed that he plans to teach niche content that may not have much valid information in ChatGPT, such as Shakespeare's early sonnets, rather than "Midsummer Night's Dreams."

Chatbots may prompt "those who prefer the original authoritative text to get out of their comfort zone and get in touch with things that don't go online," he said.

To prevent plagiarism, Aldama and other professors say they plan to set stricter standards for determining teaching expectations and grading methods. Now, it is not enough for an article to have only themes, introductions, supporting paragraphs and conclusions.

"We need to improve our level," Aldama said. "We need to infiltrate the imagination, creativity and innovative analytical skills commonly thought of as A-level papers into B-level papers."

Universities are also committed to giving students an in-depth understanding of new artificial intelligence tools. Both Buffalo University in New York and Furman University in Greenville, South Carolina, say they plan to embed discussions about artificial intelligence tools in compulsory courses such as academic integrity taught to freshmen.

"We have to add a scenario so that students can see concrete examples," said Kelly Ahuna, director of the Office of academic Integrity at the University of Buffalo. "We want to prevent things from happening, not to deal with them when they happen."

Other universities are also trying to draw boundaries for the popularity of artificial intelligence. Washington University in St. Louis and Vermont University in Burlington are revising their academic integrity policies to include generative artificial intelligence in the definition of plagiarism.

John Dyer, associate dean of admissions services and educational technology at Dallas Theological Seminary, said the language in the school's honor code feels "a little out of date anyway." He plans to revise the definition of plagiarism to cover "using text written by the generation system as his own text (for example, prompting for input into artificial intelligence tools and output for papers)."

Even so, the abuse of artificial intelligence tools is likely to never end, so some professors and universities say they plan to use testing tools to eradicate the practice. Plagiarism detection service Turnitin says it will add more features to identify artificial intelligence such as ChatGPT this year.

More than 6000 teachers from Harvard University, Yale University, Rhode Island University and other schools have also signed agreements to use GPTZero. Edward Tian, a program developer and senior at Princeton University, says GPTZero can quickly detect text generated by artificial intelligence.

Of course, some college students also see the value of using artificial intelligence tools to improve learning results. Lizzie Shackney, 27, a student at the University of Pennsylvania School of Law and Design, has begun using ChatGPT to brainstorm papers and debug coding problems.

"some subjects want you to share information rather than stand still," she said, describing her computer science and statistics courses. "the most useful thing in my brain is to understand the meaning of the code."

But she also has concerns. 'The ChatGPT sometimes misinterprets ideas and misreferences sources, 'Mr. Shackney said. The University of Pennsylvania does not have any rules on this tool, and Shackney does not want to rely on it in case the school later banned or decided that the use of ChatGPT was cheating.

Other students do not have such concerns. They shared on the forum that they had submitted papers or answered questions written by ChatGPT, and sometimes helped other students do so. TikTok's content on ChatGPT has been viewed more than 578 million times, and many people are sharing videos of writing papers and solving coding problems in ChatGPT.

A video shows a student copying a multiple-choice question and pasting it into an artificial intelligence tool. "I don't know how you do it, but I just asked Chat GPT to help me pass the final exam," he wrote. "have a good time in your study."

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.