The development of artificial intelligence systems has led to the emergence of new challenges regarding their uses, and one of the most prominent of these systems is the interactive chatbot (ChatGPT), for which OpenAI launched a beta version a few weeks ago, which raised the concerns of teachers and professors in universities and institutes as it allows students to cheat on a large scale without being able to detect them.
The ChatGPT bot differs from all available interactive chatbots in that it talks with users with high fluidity, gives written answers with a human touch to difficult questions as if you were talking to another person, and refuses to answer some inappropriate questions; Such as: how to bully or how to rob a house, and others.
ChatGPT bot and class cheat:
Due to the hype created by the ChatGPT bot, teachers began experimenting with the bot to get a sense of what they were facing. Although the bot’s answers to academic questions weren’t perfect, they were very close to what teachers would expect from their students.
So the teachers wondered, could their students use a ChatGPT bot to write reports, essays, and programming exercises without their knowledge? How will the spread of such tools affect their academic level?
Teachers usually give students the task of writing an essay because it teaches them many important skills, such as: researching the topic of the essay, categorizing and making judgments about claims, collecting the information they found, and then expressing it in their own way.
Māra Corey, an English language teacher at a Minnesota school said, “She discussed ChatGPT with her students so they could understand how using this tool could hinder their learning.”
ChatGPT beta reached 1 million users within only 5 days of its launch, some users have asked normal questions, such as: how to explain to a 6-year-old that Santa Claus is not real, or ask for new ideas for a birthday party, but Some of them went further, as there were those who asked the robot to find the problem in the code that he wrote, and there were those who asked him to write a poem and write articles on various topics.
So for some students, the ChatGPT bot has become a magical tool that does homework for them. One student told the newspaper, “He used ChatGPT twice to do his homework.” He got the idea after seeing people explaining on Twitter how powerful the tool is after it was released on November 30, 2022.
The student said, “His homework was technical terms in computer science that he didn’t know. The ChatGPT bot asked about it and got the answer almost instantly. Then it manually typed the answers into its test sheet and gave it to the teacher.
Then later that day, he used ChatGPT to help him write a piece of code for the teacher’s own homework question. The student also said he plans to use ChatGPT to cheat on exams instead of Chegg, a homework-help website he has used in the past.
The student also confirmed that he is not worried about being discovered, because he does not believe that the teacher is able to detect his answers and determine that they are the result of his use of a computer program. Therefore, teachers have found themselves facing an ethical dilemma in the use of artificial intelligence, and they must confront it.
Artificial intelligence and its moral hazards:
OpenAI developed the ChatGPT chatbot using the large language model (GPT-3.5) it had developed for years to predict language. It trained it to predict the next word in a sentence by analyzing huge amounts of internet text and finding patterns by trial and error. It can be distinguished from the human text in many cases.
The initial development of the ChatGPT bot involved AI trainers providing it with conversations in which they played both the roles of the user and the AI, assistant. It also fed answers to regular users to make its responses more conversational. Many noted its ability to produce paragraphs that were often humorous or philosophical.
However, some of his responses were completely wrong or biased, with content writers admitting that the ChatGPT bot is not perfect and sometimes gives misleading answers.
Educators assume that over time the ChatGPT bot and similar AI tools will improve and that students will learn more about them. Some say teachers will adjust their assessments to take into account the possibility of cheating, for example, they will ask students to do homework during class, where they can be watched, while others might think about how to write questions that require more thought, which is more of a challenge for the robot.
With this, greater challenges will arise for teachers. Many teachers agree that learning to write can only happen when students wrestle with ideas and put them into sentences, as students begin not to know what they want to say, and while writing they discover it.
“The writing process transforms our knowledge,” said Joshua Wilson, an associate professor in the College of Education at the University of Delaware. This will be lost if we rely on robots to write for us.
As David K. Thomson, Associate Professor of History at Sacred Hart University in Connecticut, USA, said: “The heads of departments at the university have already discussed how to deal with artificial intelligence, and they found that they must find ways to confront it.”
He added, “I discovered through my experience with the (ChatGPT) bot that it works well with the kind of questions that appear in many homework assignments, such as a question asking a student to compare the development of North and South American colonies before the economic revolution. But when asked a more complex question, such as: How did Frederick Douglass present his argument against the institution of slavery? The answer was far less convincing. Professors will have to present questions that govern analytical thinking, not just facts that can be researched, he said.
At the same time, other professors see potential gains from this technical development, as Joshua Eyler, assistant professor at the University of Mississippi, said: “This moment can be likened to the calculator moment for teaching writing, just as calculators changed the way we study mathematics, this is A similar moment to change the way writing is taught.”
While educators are considering ways to live with new AI tools, some companies are considering ways to overcome them:
Turnitin, which has developed widely used software to detect unoriginal content in scholarly papers for students and researchers to prevent plagiarism, is now looking for ways to detect material produced by artificial intelligence.
“Automated essays differ from student-written works in many ways, as students write in their own way, which is absent from ChatGPT content,” company officials say.
“OpenAI should allow the public to browse the code of a ChatGPT bot, because only then can scientists build truly powerful cheat-detection tools,” said Sasha Luccioni, a research scientist at artificial intelligence company Hugging Face.
OpenAI itself is looking for solutions:
OpenAI researcher Scott Aronson said the company is exploring different ways to fight abuse, including using watermarks or offering tools that detect ChatGPT bot answers.
Some users wondered about the feasibility of a watermark solution and whether it would be enough, to reply (Aronson), saying: “We are still conducting experiments to determine the best approach to fight abuse and there may be more than one approach together.”
While the ChatGPT bot had its own ideas for a solution, when asked how to counter the possibility of cheating, the bot made several suggestions, including educating students about the consequences of cheating, providing monitoring while solving tests, making questions more complex, or giving students The support they need so they don’t see the need to cheat.
Finally, the bot explained that it is important for the teacher to communicate clearly to students about their expectations of academic integrity and to take steps to prevent cheating, as this step will help them create a culture of honesty and integrity in their classroom.
Why do we use artificial intelligence despite its ethical dilemmas?
The technological revolution is changing our lives at a tremendous speed, as it greatly changes the ways in which we work and learn, as well as the ones in which we live together, and there is no doubt that artificial intelligence is going through a huge stage of development during this current period, as many of its new applications and tools have appeared in many sectors, including security, environment, research, education, health, art, and trade.
Artificial intelligence is expected to change many professions that humans do soon, so the right approach in facing these technologies is to learn how to use them to improve humanity.
Artificial intelligence will lead to a new form of human civilization, so we must ensure that it is developed through a humane approach based on values because the revolution of artificial intelligence opens up new horizons for human civilization, but the ethical dilemmas that arise for it require careful study.