Alarmed by AI chatbots, universities are beginning to revamp the way they teach

While grading essays for his world religions class last month, Antony Aumann, a philosophy professor at Northern Michigan University, read what he called “the best paper in class.” He explored the morality of burqa bans with clear paragraphs, appropriate examples, and rigorous arguments.

A red flag immediately went up.

Mr. Aumann asked his student if he had written the essay himself. The student confessed to using ChatGPT, a chatbot that provides information, explains concepts and generates ideas in simple sentences – and, in this case, had written the document.

Alarmed by his discovery, Mr. Aumann decided to transform essay writing for his classes this semester. He plans to require students to write first drafts in class, using browsers that monitor and restrict computer activity. In later versions, students must explain each revision. Aumann, who may forego essays in subsequent semesters, also plans to embed ChatGPT into classes by asking students to rate the chatbot’s responses.

“What happens in class is no longer going to be, ‘Here are some questions – let’s talk about it among us human beings,’” he said, but instead “it’s like, ‘What- what this alien robot thinks too?’”

Across the country, college professors like Mr. Aumann, department chairs, and administrators are beginning to rearrange classrooms in response to ChatGPT, sparking a potentially huge shift in teaching and learning. Some professors are completely redesigning their courses, making changes that include more oral exams, group work, and handwritten assessments instead of typed assessments.

The movements are part of a real-time struggle with a new wave of technology known as generative artificial intelligence. ChatGPT, which was released in November by artificial intelligence lab OpenAI, is at the forefront of change. The chatbot generates oddly articulated and nuanced text in response to short prompts, with people using it to write love letters, poetry, fan fiction — and their schoolwork.

This has upset some middle and high schools, with teachers and administrators trying to discern whether students are using the chatbot for homework. Some public school systems, including New York and Seattle, have since banned the tool on school Wi-Fi networks and devices to prevent cheating, though students can easily find workarounds to access ChatGPT.

In higher education, colleges and universities have been reluctant to ban the AI ​​tool because administrators doubt the move will be effective and they don’t want to infringe on academic freedom. This means that the way people teach changes instead.

“We try to institute general policies that definitely support the faculty member’s authority to lead a class,” rather than targeting specific methods of cheating, said University of Florida provost Joe Glover. “This won’t be the last innovation we have to deal with.”

This is all the more true since generative AI is in its infancy. OpenAI should soon release another tool, GPT-4, which generates text better than previous versions. Google has built LaMDA, a rival chatbot, and Microsoft is discussing a $10 billion investment in OpenAI. Silicon Valley startups including Stability AI and Character.AI are also working on generative AI tools.

An OpenAI spokeswoman said the lab recognizes that its programs can be used to trick people and is developing technology to help people identify text generated by ChatGPT.

In many universities, ChatGPT is now at the top of the agenda. Administrators are creating working groups and holding university-wide discussions to respond to the tool, with much of the advice being to adapt to the technology.

At schools such as George Washington University in Washington, DC, Rutgers University in New Brunswick, NJ, and Appalachian State University in Boone, North Carolina, professors are phasing out homework assignments at home, open book – which have become a mainstream assessment method in the pandemic but now seem vulnerable to chatbots. Instead, they opt for class assignments, handwritten assignments, group assignments, and oral exams.

Gone are the prompts like “write five pages on this or that”. Some professors instead craft questions that they hope are too clever for chatbots and ask students to write about their own lives and current events.

Students “plagiarise this because assignments can be plagiarized,” said Sid Dobrin, chair of the University of Florida’s English department.

Frederick Luis Aldama, chair of humanities at the University of Texas at Austin, said he plans to teach newer or more specialized texts that ChatGPT might have less information about, such as early sonnets by William Shakespeare instead of “A Midsummer Night’s Dream”.

The chatbot can motivate “people who lean into canonical primary texts to go beyond their comfort zone for things that aren’t online,” he said.

In case the changes aren’t enough to prevent plagiarism, Mr Aldama and other professors said they plan to institute stricter standards for what they expect of students and how they grade. It is no longer enough for an essay to just have a thesis, an introduction, supporting paragraphs, and a conclusion.

“We have to improve our game,” Mr Aldama said. “The imagination, creativity and innovation of the analysis that we generally consider to be an A item must carry over to the B range items.”

Universities also aim to educate students on new AI tools. The University at Buffalo in New York and Furman University in Greenville, SC, said they plan to incorporate a discussion of AI tools into required courses that teach entry-level or first-year students concepts. such as academic integrity.

“We need to add a storyline to this, so students can see a real-world example,” said Kelly Ahuna, who heads the University at Buffalo’s Office of Academic Integrity. “We want to prevent things from happening instead of catching them when they happen.”

Other universities are trying to draw boundaries for AI Washington University in St. Louis and the University of Vermont in Burlington are drafting revisions to their academic integrity policies so that their definitions of plagiarism include AI generative

John Dyer, vice president of enrollment services and educational technology at Dallas Theological Seminary, said the language of his seminary’s honor code seemed “a little archaic anyway.” It plans to update its definition of plagiarism to include: “the use of text written by a generation system such as its own (for example, entering a prompt into an artificial intelligence tool and using the output in a article).”

The misuse of AI tools is unlikely to stop, which is why some professors and universities have said they plan to use detectors to eradicate this activity. Plagiarism detection service Turnitin said it will integrate more features to identify AI, including ChatGPT, this year.

More than 6,000 teachers from Harvard University, Yale University, University of Rhode Island and others have also signed up to use GPTZero, a program that promises to quickly detect text generated by ‘IA, said Edward Tian, ​​its creator and senior at Princeton University. .

Some students see value in adopting AI tools for learning. Lizzie Shackney, 27, a student at the University of Pennsylvania Law School and School of Design, started using ChatGPT to brainstorm articles and debug sets of coding problems.

“There are disciplines that want you to share and don’t want you to spin your wheels,” she said, describing her courses in computer science and statistics. “The place where my brain is useful is figuring out what the code means.”

But she has scruples. ChatGPT, Ms. Shackney said, sometimes incorrectly explains ideas and miscites sources. The University of Pennsylvania also hasn’t instituted regulations regarding the tool, so it doesn’t want to rely on it in case the school bans it or considers it cheating, she said. declared.

Other students have no such qualms, sharing on forums like Reddit that they’ve submitted assignments written and solved by ChatGPT — and sometimes for other students as well. On TikTok, the hashtag #chatgpt has over 578 million views, with people sharing videos of the tool writing articles and solving coding issues.

A video shows a student copying a multiple-choice exam and pasting it into the tool with the caption saying, “I don’t know about all of you, but I’m just asking Chat GPT to pass my final. you to study.