ChatGPT slipped yesterday in the Spanish language and literature exam of the entrance exams to the university in Catalonia as one more student. And how has your performance been? Have you passed the exam? What grade would the examiners have given him if he had been a person?
This chatbot has been able to carry out the evaluation in just a few minutes without having to go to a university classroom, without having to prepare for the exam and without facing the nervous breakdown that students experience when facing the first selectivity exam.
The teacher Montserrat Torra, a Spanish language and literature teacher for high school students, already retired, has accepted the proposal of this newspaper to correct the exam carried out by artificial intelligence following the established evaluation criteria.
A first impression is that ChatGPT has managed to succeed in reading comprehension exercises. Content correct and well adjusted to what was requested in the statement.
However, it seems that the chatbot has never read the Madrid writer Antonio Buero Vallejo or the Catalan author of Nada, Carmen Laforet.
Although far from leaving blank the questions that were not known and that some prudent students in their position might not have answered, artificial intelligence has chosen to “invent” the answers, perhaps as some other clever student would do. If it strains, strain. But, indeed, the answers have not been valid. See example below.
As for the written expression, the conversational assistant has not had any problems to function. “Option A has been answered without making any errors, while in option B there was a small mistake,” says Professor Torra. Artificial intelligence has had trouble distinguishing between part and part. But other than that, it has gotten a good rating.
In short, says Torra, “Reading and writing, a 10! And the occasional mistake in grammatical knowledge,” says the teacher.
Adding together the responses of reading comprehension, written expression and linguistic reflection of the exam, the IA student has obtained a grade of 8 out of 10. Both in option A and B.
