The ‘dean’ of Nova SBE says, in an interview with Lusa, that ChatGPT is a “challenge” for universities and “questions many professions”, from teachers to journalists, with these new artificial intelligence (AI) solutions.
Pedro Oliveira, who has been director of Nova School of Business & Economics (Nova SBE) since January 19, considers that ChatGPT “is in fact causing a great impact in many industries”, which includes education, training.
This is because “suddenly there is a communication system with which it is possible to maintain a conversation in writing, to whom we ask questions and which typically gives us answers faster and even with, let’s say, on average, higher quality than that, for example, that people give us when we submit them to an exam or when we try to talk to them”, says the director of Nova SBE.
For anyone interested in technology, this ‘chatbot’ “is an incredible development, but it raises so many questions” and “so many questions”, because “it changes the way we do so many things”, he says.
“Today we are talking about artificial intelligence associated with reading magnetic resonance images, but also autonomous driving of cars. After all, maybe soon we can even go home without having to worry about driving the car”, he predicts.
ChatGPT opens up “such a wide range of possibilities that, from artists to musicians, teachers, journalists, radiologists, I think we all have some reason to be a little anxious about what will happen”, anticipates Pedro Oliveira.
In this case, artificial intelligence “has arrived at the production of responses”, this system “allows us to have a very intelligent conversation” and “there are those who say that it is much more intelligent to have conversations with ChatGPT than with some people and get interesting answers and well founded, sometimes with some mistakes”, says Pedro Oliveira.
In a “short time” the ChatGPT 3.5 version gained prominence – and 4.0 is even announced soon – which “means that there is, in fact, a revolution coming that calls into question many professions like ours, from trainers to and people who, for example, have to assess students’ knowledge, making it increasingly difficult to determine whether that knowledge is original or whether it was produced by a machine”, warns Nova SBE’s director.
“Perhaps for universities it is the main challenge”, but in other industries there are also challenges, he admits.
At universities “we already had some technology to detect whether that information was plagiarized”, but now “it becomes much more difficult” because “the narrative that is produced by ChatGPT is similar, it is inspired by human narrative”, he explains.
“We now know that it is possible to answer exam questions through ChatGPT”, he warns, giving the example of the operations management course, in which at the Wharton Business School, at the University of Pennsylvania, “the computer was given the exam that would normally be given to students”.
And the “computer passed, that is, the ChatGPT gave the right answers”, as it also passed in accessing the legal profession in the United States, he pointed out.
“We started to realize that it is very scary, because in fact it is a new challenge for us”, because if you have to correct an exam it is not easy to know who did it: the student or the machine, which raises challenges such as how it will be done the assessment from now on, comments.
In Denmark, where he has worked and lived, “the recommendation is that exams become more and more oral-based, rather than all written”, he says.
“If we have the ability to give our students oral exams, then we realize that it is the person in front of us who is transmitting that to us in the background”, he argues.
And how is this resolved in Portugal? “There isn’t a very complete answer to that yet,” he says.
One of the paths of universities “is to deny, to prohibit”, this is the temptation “when we are confronted with an innovation, with a different idea, which can be disruptive of our ‘modus operandi’”, he continues.
At her daughter’s secondary school in Denmark, where students started submitting work co-produced by ChatGPT, “the immediate response was – and understandable in a way – ban”, she says.
Now, “we cannot always prohibit access to new technologies”, he argues, noting that what is happening at the moment is not very different from the advent of the Internet and search engines.
“There are mixed feelings here: on the one hand, we want to embrace all these new technologies, on the other hand, they are almost frightening, therefore, today there is already talk of the possibility of ChatGPT being a co-author” of academic ‘papers’, advances the ‘ dean’.
“There are those people who are still almost in denial and today, when they write an article for a newspaper, they write underneath saying that this article was produced without any use of ChatGPT. What I think is not sustainable in the long term, because what makes sense is to do – I think I’ve done this myself – write articles and say this article was written partially or or at least with ChatGPT support, because there is no bad about it”, argues Pedro Oliveira.
Incidentally, “today it is completely socially acceptable that before I write an article I do some research ‘online’ to understand any topic and, therefore, if I do this with the search engines, I will certainly be able and even should be able to do this in the short term. term with other types of technologies, namely artificial intelligence”, he considers.
With ChaGPT, which so far is available free of charge, it is possible to have an “intelligent” conversation, with whom you can “learn” or “study” and that “is quite different from what we were used to”, he stresses.
The “fascination” with ChaGPT comes “from the fact that, all of a sudden, the machine really starts to replace the man”, considers Pedro Oliveira.
This “is no longer new” because machines were already replacing men in “lesser added value activities”, such as in factories, but today “it is not like that”.
“They are already professions for more sophisticated individuals who have had to study for more years. It’s doctors now who are being (…) challenged in relation to ChatGPT, professions like university professors, journalists, who suddenly can see ChatGPT writing their own pieces, their own articles, artists, because suddenly it’s possible make poems”, emphasizes Pedro Oliveira.
In general, “almost all professions (…) have some associated risk”, he says, giving an example of what happens to radiologists.
“The best way to read a medical image produced by an MRI is already a machine, not a doctor. It is not the radiologist, necessarily, who gives me the best result, because the machine has the ability to learn. Perhaps the machine has already read millions of images (…) and realized which were the malignant and benign tumors, he points out.
What was previously done “by people, by doctors who were there, patiently, analyzing the images, today the same answer is given to me by a machine in a much more reliable way” and then “the machine already replaces the same man and now to the radiologist”, he underlines.
The European Commission “has already approved the use of this type of readings to inform patients, in particular when the cases are not very complex”. In view of this, the question that remains in the air is: “Are we really all going to be replaced by machines?”.
This post is also available in: Português