Bacon!

Our language does the same: it simplifies the world in order to make it understandable to us, so that we can isolate phenomena and objects, experience them as separate, and use and talk about them in useful ways. That is not a problem. The problem arises when we forget this, when we forget what language is.

In other words, we should remember that something falls through the cracks when we categorize the world and create words by delimiting the phenomena of reality. This something continues to exist in the cracks.

Babies see this something — language-less as they are, they can be fascinated by a part of a wall that has visible contrasts in colors their eyes can perceive; they are entertained by a version of the world that, taken as a whole, has no word for it. They do not concern themselves with observing only what we delimit as objects. Objects are given names so that we can later think of them as precisely objects.

To distinguish and name something is useful for us — language is utility, a tool. Everything we name is named on the basis of human needs in the world; language is therefore not another form of identical reproduction of everything that exists.

In the 17th century, nature was frightening, dark, and feminine (Survival of the Richest, p. 59). It had to be tamed. It was counted and categorized.

All of this, this way of seeing, is reproduced by artificial intelligence. It is useful. But AI is built on an incomplete and therefore shaky foundation; namely, humanity's quantification of the world and the resulting, though useful, constructed reality. Thus AI has limitations: it leaves out what cannot be counted or described with language (because, as we remember, language is always and necessarily incomplete). The machine can only use data; it does not look into the cracks, for what is there is not necessarily accessible to us. Yet this hidden thing exists, and it affects us.

Poetry tries to express what we do not have words for, but sense intuitively. In this way, literature is an acknowledgment that we cannot observe and name the whole of complex reality, or every part of it. But we live in a culture that often tells us the opposite.

So, the moment something is delimited and categorized, we have made a linguistic simplification. This is of course useful for us humans, but we must not forget that it is artificial — a construction, not reality itself. Machines are fundamentally built on this artificiality. They are trapped in their schematic "nature." They can do what humans can do when we use language, but not more — albeit much more efficiently.

The Enlightenment gave us much. Let us give the future more.