The extinction of the language is the catastrophe from which not even ChatGPT will save us

One of the most important moral philosophy books of the last century, After virtue by Alasdair MacIntyre, begins as if it were a science fiction novel. It is said of men who, following an unspecified “catastrophe”, have lost the sense of the culture in which they live. As rubble, some words remained, evaluative ethical terms such as good, bad, right, unjust or certain expressions with which the survivors indicate to their fellow men what they should do in certain circumstances. But there is no longer any trace of the conception of man from which these terms derived their meaning; the historical-cultural context has disappeared within which human life still appeared as the life of an ego, whose biography could be told as a whole. And so, one might add, we have become “one, none, and a hundred thousand.”

I have often reflected on this catastrophe, wondering among other things if we would ever get out of it. But for some time now I have changed register; all I can do is prefigure its possible final outcome: the extinction of the language. Important signs in this direction can be seen everywhere: in the little attention that schools of all levels pay to the most elementary rules of grammar; in the difficulty of the younger generations to put even the most elementary thoughts in writing; in the impressive number of hours they spend online every day; in the poverty of their lexicon, often reduced to a few dozen words, whose semantic depth does not go beyond that of a road sign; in the lack of interest in reading in general; in the poor quality of what passes for literature; in the neglect of certain newspaper articles (in recent days a well-known newspaper wrote on the front page: “Poker from Milan to Maradona with two goals from Leao, Diaz and Saelemaekers”. Poker rightly suggests that Milan scored four goals, but the syntax of what following misleads one into thinking that he did six).

And all this while the new artificial intelligence models, see ChatGPT or GPT-4, are capable of writing an article, of answering our questions in writing with an already good quality of writing, which will soon become impeccable. Very well, some will say, writing will not die, machines will keep it there. Yes, but at the same time they could also make it even more useless.

Given that I am hyperbolizing a catastrophe which, like all catastrophes announced by men, will end up being only presumed, certainly the impoverishment of our language is real and produces such profound inequalities that artificial intelligence will hardly compensate for them. To tell the truth, one could even imagine a situation in which we will all have the possibility of having a machine suggest the right words to use at the right time. Maybe the machine will even be intelligent enough to adapt them to each of us, making us feel unique. But who tells us that in the long run this uniqueness will not turn into ever greater isolation? And what will all this mean for those who have never been used to hearing the sound and depth of words? Will it be able to grow thanks to the machine? Or, as I think, will the machine only accentuate his poverty and linguistic passivity?

As for isolation, yesterday all citizens knew more or less what the others knew. For example, speaking of public opinion meant referring to a common area. But algorithms, as is known, make mass personalization possible, a generalized individualization that no longer has anything to do with all this. Having access to information does not mean having access to a shared world, but simply enlarging our individualized world. Therefore, the risk is high that in the end, instead of being unique, one will feel isolated and massified: “One, no one and one hundred thousand”.

Lacking a common point of reference, we no longer know what others presumably know or don’t know, nor are we able to evaluate our own ignorance. On the other hand, as in ancient magical divination, the language of our machines, the machine learning algorithms, mostly obscure to our minds, are used to “manage” and “predict” phenomena, not to understand them. Thus we will adapt to the intelligence of machines (yes, this is possible!), without even knowing that the latter have now definitively set aside the dream of emulating ours because it is too complicated, beautiful and mysterious. And to think that this awareness would be enough to keep the astonishing sense of our language awake, even making us hope that we could be the ones to creatively use that of the machines. Instead we think we’re solving the problem by banning the use of ChatGPT: a decision so devoid of human intelligence that it seems to have been suggested by ChatGPT itself.