Mastering Semanticization and (Re)Ontologization Skills: The Key to Excelling with AI Interaction and Reversing Cultural Decline

In a previous article (see link 1 below), I argued that among the most important skills in the era of AI are elements that are far from technical, such as the ability to semantize, or assign a coherent meaning to an object and describe it in a fully comprehensible manner, and the ability to (re)ontologize, meaning the skill to grant an element the status of existence, extracting and differentiating it from the context in which it was immersed. These skills might seem philosophical at first glance, but, after analysis, they are considered more relevant than others, like the usual “problem-solving” or “effective communication”, which generically sound always correct in any context.

Following that article, the exercise I attempted was to understand how these two characteristics could be practiced and developed. After all, if they are important, it is equally important to possess and improve them.

By making more pragmatic reflections, I've come to some conclusions that can both answer what was sought and expand the reasoning towards other perspectives too.

?

Semanticization.

To understand how to improve this skill, I began by asking myself a simple question: how many words does a person typically know? The answer was enlightening in itself.

I found that the Treccani Encyclopedic Institute, one of the most important of this kind in Italy, covers about 450.000 words in the Italian language. Of these, approximately 10% are part of the lexicon commonly used by those with a medium to high level of education and, of these, about 15% are used as a basic vocabulary with which we cover almost 98% of our conversations (see link 2 below).

Translated into more direct terms, to talk about almost everything, we only use 1-2% of the words potentially available to us.

It follows that if we were to exclude all the terms used in highly technical contexts, we would still have at our disposal a vast amount of words that we typically don't use, but with which we could describe the world around us more accurately. Just as we could communicate our thoughts in a more sophisticated manner.

Not only that. With a richer lexicon, we would also be able to distinguish more subtle nuances of both what we observe and what we think, simply because we would associate more appropriate words with qualities that otherwise we would not characterize precisely enough. In summary, with a broader vocabulary, we can produce better ideas and tell them more effectively.

It goes without saying that expanding our lexicon could come from studying a dictionary, although, from my point of view, this would be an extremely tedious activity.

To avoid this, through some online research I have discovered websites run by enthusiasts (see link 3 below as an example) that offer, through newsletters, a single word analyzed each day. This includes a precise description of its meaning, etymology, and examples of use in different contexts to help grasp the differences and appreciate the nuances of its application, making the learning process significantly more digestible and interesting.

Though it may seem at first glance like a school exercise, the truth (backed by the numbers mentioned above) is that it leads to unexpected discoveries, among which:

• We learn words whose existence we were completely unaware of but which are perfectly applicable to everyday contexts, thereby increasing the repertoire of words at our disposal for describing and communicating more accurately.
• We discover that sometimes the meaning we attributed to a word is not the correct one, thereby increasing our precision in using the language.

If one were wondering at this point what this has to do with artificial intelligence, the answer is quite simple: the ability to appropriately describe a context in relation to a specific object, and to formulate coherent requests about it in a conversational manner, is exactly what is needed to interact effectively with a Generative AI tool (e.g., ChatGPT, Google Gemini, Anthropic Claude, and similar). The more we improve our accuracy in the language with which we interact with automation, the more we can achieve what we actually expect from these tools.

Beyond specific technicalities (e.g., Zero-Shot Prompting, Chain-of-Thought Prompting – see link 4 below), the capacity for dialogue is thus the most basic element of Prompt Engineering, and therefore needs to be developed.

Two notes regarding this:

• Beyond interaction with machines, it's essential to emphasize the importance of this also for relationships between people and for gaining a deeper understanding of each other.
• In addition to the two points mentioned earlier, there is a third one: enriching one's lexicon and understanding the etymology of words adds historical depth to the language and, therefore, to ourselves and our society. This will be particularly relevant for discussions that will take place later on.

?

(Re)ontologization

Generative AI tools are not merely an “additional dropdown feature” in the toolbar of a productivity application (e.g., MS Word). They represent a radical paradigm shift through which one can have access to a (nearly) general-purpose personal assistant, with whom to interact in a new way. The availability of innovative technology can, consequently, enable equally innovative use cases.

The identification of these cases, however, comes from a rational analysis of current systems (e.g., business processes or products) and an understanding of how they can be revolutionized thanks to technology.

This is the activity of anticipated (re)ontologization: identifying innovative elements and their coherent integration within a reference context.

It is clear at this point that this concept, which seems so philosophical, can be exercised in an extremely pragmatic manner within a business context through at least two well-known activities:

• Process (re)engineering – namely, the systematic study of how specific activities can be optimally performed, often by incorporating innovative technologies. This involves decomposing complex processes into logical blocks, understanding, revising, and recombining them. These steps are essential for granting a new element the status of existence (re-ontologization), which, in this case, practically translates into the redefinition of new ways of working.
• Prototyping – similar to the reengineering of processes, the rapid realization of ideas (e.g., Minimum Viable Product), their quantitative analysis, and their continuous evolution, constitute activities of redefining elements that did not exist before but are now gaining a life of their own.

These are decidedly concrete and useful activities for a company, which can be carried out (and probably already are) in any function, but on which there is now the opportunity to extract even more value if emphasis is placed on the underlying theoretical concept: the focus on bringing to life something new that did not exist before.

Again, what does all this have to do with AI?

A systematically analytical approach to reality, aimed at understanding and proactively manipulating it for purposes such as innovation, allows for the development of clearer ideas about the systems around us and our goals for them. This approach is not just theoretical; it is another (if not the most) fundamental element for Prompt Engineering.

In fact, everyone has likely experienced interactions with others and realized that the discussed ideas were not perfectly clear, their boundaries were not well defined, nor was the actual goal they intended to achieve. When applied to interactions with machines, this leads to making requests that are not always focused, resulting in responses that are not entirely useful for our purposes.

However, if one were to have greater mental clarity and a clearer understanding of the context and objectives, then it would be possible to specify our needs more precisely. By then leveraging the advanced semantic capabilities discussed earlier, we would pose the right questions in an appropriate manner to Gen.AI tools, obtaining exactly what we want and maximizing the benefits of their use.

?

As it becomes clear, the analysis underscores elements that are hardly technical and, indeed, relate to very human skills, such as developing analytical minds and precise communication. These elements are not only fundamental for interacting with advanced automation but are also at the core of a person's human and professional solidity, which is therefore worth cultivating and practicing.

?

To conclude, all this reasoning also brings me back to an article I recently read in the March 2024 issue of Limes, an Italian geopolitical magazine. In a summary of one article, focused on the contrast between “TINA” (i.e., There Is No Alternative) and “TARA” (i.e., There Are Reasonable Alternatives), the author states: “The crisis of Western democracy is evidenced by the triumph of no-alternative theses, or the rejection of debate. Denial of time and history [reference here to the historical depth mentioned above], this arrogance signals a remarkable cultural decline” (see link 5 below).

It is clear that the aforementioned cultural decline has passed, and continues to pass, through a prolonged devaluation of sophisticated intellectual exercise, as well as a preference for light entertainment over the greater effort required by in-depth analytical study (see the success of social media). The degeneration towards "TINA" rather than "TARA" (including the "rejection of debate") also stems from an increasingly poor understanding of the context in which we find ourselves and from the inability to argue alternative ideas (referring, respectively, to the (re)ontologization and semanticization mentioned above).

Starting over by investing from the ground up, with renewed attention to building solid analytical minds capable of arguing effectively and precisely, could therefore not only prepare workers and future generations to interact positively with AI, but also to rebuild the elements that counter the mentioned cultural decline. In this, businesses and governments clearly play a crucial role.

I leave here below the links to some additional article that you may find interesting:

• A philosophical approach to strategy leads to the definition of the most important upcoming competencies for companies
• Treccani, lessico
• Una parola al giorno
• Prompt Engineering Guide
• Limes, Rivista Italiana di Geopolitica