ChatGPT takes typing as a mechanical task – 02/25/2023 – Thaís Nicoleti

Latest news in the field of technology, the ChatGPT application has impressed the public for its ability to write texts “as if it were a human being”. In general lines, the system learned to write with the immense amount of text files stored in its “brain”. Is written language being relegated to the background, with text production likely to be delegated to a machine?

The application, it is said, is capable of making summaries and paraphrases and, as it seems, unlike most human speakers of Portuguese, it has no difficulty correctly applying the indicative accent of crasis. Technologies, as we know, tend to meet needs – and this case will not be different. In general, they replace tasks that take up time and, because they are mechanical, can be delegated. The bus collector, which was needed when passengers paid in cash, is easily replaced with a card reading system. Is writing texts, then, a “mechanical” task, which can be performed by an artificial system?

Apparently, the answer is affirmative, as large sums of money would not be invested in the development of a useless project. Now, artificial intelligence manages to do in seconds what we humans would take a long time to achieve without the certainty of getting good results. What’s the word that doesn’t come to mind? What is the more accurate term? The robot solves the problem even before we ask the question.

Students may have been excited to hire this helper to do their homework, but if they do, obviously, the homework will have lost its raison d’etre. Why do teachers ask for summaries, paraphrases and writing exercises? All academic life has been built by the practice of reading and producing texts. It is understood that the capacity for verbal expression accompanies intellectual maturity. The more you read, the more you learn to write. Reading provokes our thoughts, and text is a way of organizing them. Organized, they constitute knowledge.

It is worth asking whether we really want to delegate to robots the task of organizing our ideas, as if this organization were something alien to us. Do we want them to express themselves in our place, with the words that some algorithm chooses? Behind a robot, there is a programmer, to whom, ultimately, we will be entrusting, as if it were a minor thing, the expression of our thoughts. The most important issue, however, is perhaps not even that, as “fighting” with technology has never brought good results. This arrives, in general, as a result of some demand.

It is not new that, thanks to technology, we know that not even newspaper texts, which are generally short, are read to the end. A lot of people are content with the title and forward an article or news that they haven’t read to their friends on their cell phone and that they may read at some later date or that they may never read. There are others who venture as far as the first paragraph. How many will have made it this far? If we don’t have time to read, what about writing, an even more complex task? Still, we write.

Search engines today are capable of detecting relevant information in texts for those looking for them. Of course, whoever writes should help them, using the simplest and most objective language possible so that the robot – which, after all, is just a robot – can find them and offer them to the querent. In other words, search robots are already the mediators between the author and the reader. Of course, reading is much more than “looking for information” in a text, but it can also be just that. Purely informative texts, it seems, can indeed be written by ChatGPT, which will know better than any human being the ideal way to communicate with search robots.

The field of language, however, although it contains a lot of objective data, capable of learning by mathematical systems of artificial intelligence, it will hardly be dominated, in its entirety, by robots. In the communicative process, we do not always realize how many things we do not say explicitly, but are presupposed or implied in the context. We humans, when we say, for example, that someone has stopped smoking, we don’t need to explain with words that the person was a smoker, because this is presupposed in the verbal form “to stop” in this context. Likewise, we make countless assumptions based on evidence, which do not need to be explained at the foot of the letter, as they are fully understandable by this mysterious natural intelligence of ours, which at all times relates what is said at the moment with what has already been said. before somewhere else, which the application cannot do.

We are also capable of great linguistic antics, such as expressing our thoughts through paradoxes, which are apparently absurd (for example, the “discontented contentment” with which Camões defined love) and even through irony, which is the prodigy of saying a thing meaning its opposite, or by preterition, which consists of saying that you are not saying what you are actually saying (“I won’t call you rude because it would be impolite”, for example).

Even considering that the robot is not capable of mastering all the resources of verbal language, its arrival may be a sign that a good part of the written texts are repetitive and merely informative, not being, therefore, individual expressions. Moreover, we cannot ignore that, thanks to technology, oral communication has gained considerable space. Anyone who searches for anything on the internet will find videos, from one that teaches how to install a clothesline to one that explains Hegel’s philosophy. Nothing prevents us, by the way, from watching them in sequence, because, in our fragmented way of living, one and the other can have the same importance.

Schools, perhaps in the near future, will no longer ask students to write texts, which, after all, are already tasks that robots can perform, although we must bear in mind that machines are only able to “create” what already exists, that is Yes, it manages to organize ideas, but it doesn’t actually create anything. Even so, the issue of plagiarism, a major concern of an ethical nature in academia, takes on new contours. Would using the app to paraphrase, for example, be ethically acceptable?

Everything leads to believe that the ability to express oneself orally will become required of students, as, in fact, much was done in the past – at least, until a robot is invented capable of speaking for us.