The advent of AI ChatGPT has turned the technology debate upside down. Artificial intelligence (AI) experts and users test their abilities to answer questions and follow conversations, share their most original texts on social networks and imagine what their potential could be. But while this is happening, several educational institutions around the world have seen the danger. Schools in New York, Los Angeles or eight major Australian universities have decided to veto their use by students, reopening an old controversy over technology in the classroom.
These institutions saw in ChatGPT the ultimate student trap. A machine that can answer any question that appears in a textbook in seconds, prepare an essay on Cervantes in the format of a rap song, or solve advanced algebra problems. “While the tool can provide quick and easy answers to questions, it does not promote critical thinking or problem-solving skills necessary for academic and life success,” said Jenna Lyle, a spokeswoman for the New York City Department of Education. The district with the most students in the entire US.
His organization acknowledges that the new artificial intelligence poses “concerns about negative effects on student learning, as well as the security and accuracy of content,” it added in a statement explaining the decision to veto its use. in the centers. Their Los Angeles counterparts made the same decision on November 12, just one day after the launch of this AI.
These warnings highlight two major problems that ChatGPT currently has. The first is that sometimes it is wrong. It can give wrong answers, but because it does so with high-level reasoning, it’s very hard to spot its mistakes unless you know in advance what the machine is writing about. The second is that he doesn’t disclose the sources of information he uses for his answers, making it difficult to contrast his statements or drill down on points of interest to the user.
From OpenAI, the foundation that created it, they do not hide these limitations and remember that this is a testing system. “ChatGPT is incredibly limited, but it’s good enough at a few things to give a false impression of greatness,” Sam Altman, its CEO, admitted in December. “It is a mistake to trust him with something important at this time. This is a preview of our progress, but we still have a lot of work to do in terms of robustness and accuracy,” added the head of OpenAI, an NGO. A for-profit fund co-founded by Elon Musk in 2015 to develop artificial intelligence systems.
Eight major Australian universities have also decided to act after discovering some students were completing academic work with AI answers. His solution is to “predict” the spread of ChatGPT usage among students by putting pen and paper back.
“Our universities have revised how they deliver assessments in 2023, including proctored exams, increased use of exams and paper-and-pencil tests,” said Matthew Brown, deputy director of the eight-institute co-ordinating group: “Assessment redesign is critical and this work continues For our universities, as we try to stay ahead of the development of artificial intelligence. “
elDiario.es asked the Spanish Ministry of Education for its position on ChatGPT. Sources in the department, headed by Pilar Alegría, explain that it is “analyzing and studying the tool and its results” with the National Institute of Educational Technology and Teacher Training.
It seems clear that ChatGPT may have an impact in the classroom that hasn’t been seen since the Wikipedia schism. “At the educational level, some suggest that one of the outcomes could be the end of homework as we traditionally know it, because the tool can help students with homework. Others suggest that this help actually came from the family or the academy (if they could afford it) And now it will be technology that will help us”, summarizes María del Mar Sánchez, researcher specializing in educational technology and professor of didactics and school organization at the University of Murcia.
“In any case, it’s becoming clear that the traditional model of homework needs to change,” reflects the expert: “I think what these technologies are doing is inevitably putting the need to consider new educational strategies on the table.” The education system reflects how to integrate these tools, which will be more and more common in society. effective way.”
Sánchez notes that the introduction of new technologies in the classroom should take into account the age of students and know the fundamental pedagogical principles for it, not forgetting “addiction and the risks of its abuse”. “But banning technology, as is happening in some places, does not make sense in the digital world because we are denying the opportunity for training in a key aspect of today’s society that is learning to live healthily with technology. Build an adequate digital identity and overcome the digital divide,” he said in a conversation with this media.
Manuela Battaglini, a lawyer and researcher who specializes in digital ethics, recalls that the debate about technology, which allegedly intends to destroy students’ critical thinking, is recurring. “This technology, and many others throughout history like Google Search, remind us (and we seem to ignore them) that learning is not a category that resides only in the reasonable quadrant, it is much more. This is the connection. many categories. This is where ChatGPT and the problems of education and plagiarism and zero effort on the part of students come into play. But wasn’t plagiarism in education characteristic of him? What changes is the technology that enables it. says
“Noam Chomsky wonders: What are we learning from this technology about knowledge, reasoning, concrete language, a basic component of human cognition? It turns out that there are many flaws in these models that can be fixed. More data, more parameters, more intelligent programming, but there are simple and fundamental An imperfection that can never be remedied by these methods is instead magnified by them: they tell us nothing about language, thought, the basic foundations of human cognition, or what it means to be human,” says Battaglini.
“I think the situation tells us to focus on what these models cannot do in the education sector, including many others, and start innovating and changing its parameters,” he concludes.
In any case, controversy may have a shorter path than others, such as using Wikipedia. After the backlash from the New York schools, OpenAI decided to work on a feature so that its artificial intelligence can detect when a text was written by it. “We don’t want ChatGPT to be used for deceptive purposes in schools or anywhere else, so we’re already developing mitigations to help someone identify text generated by ChatGPT,” a spokesperson told The Washington Post.
The mentioned, in turn, does not take sides. “ChatGPT is a language tool developed by OpenAl and its use in education will depend on the context and the way it is implemented. It can be useful to help students better understand certain topics and concepts, but it can also be misused. It is important to remember that ChatGPT is “Only the tool and its use will depend on the user and the context in which it is used,” he says when asked about the implications for education.
Source: El Diario
