This is a new series working out the new era of machine learning. Although I aim to provide a complete set of tutorials (incl. pragmatic programming how-tos), I will be turning all the stones and correcting any terminology which is, in passing, has been assumed and accepted without much thought in contemporary literature on the topic. In the first part, I will mostly dissect the genealogy of the common terminology and propose an alternative.

I am mostly opposed to calling any machine learning or programmed environment Artificial Intelligence. First, calling it artificial is a byproduct of our congenital defect of seeing the world and its representations from our perspective. From this anthropocentric position, the Human is the center of the universe, and everything else becomes a mere shadow. Moreover, machines (or humans) do not produce intelligence in the learning process; we merely become accustomed to the forces around us and adapt to the active forces, or in the best case, we can convert the reactive forces to active forces*.

Intelligence is often seen as a measurable trait, the sum of one’s ability to learn, adapt, and solve problems. But this rigid, quantitative approach reduces intelligence to something mechanical when, in reality, it’s far more fluid and elusive. Intelligence is not just about problem-solving; it’s about creating meaning, suffering, interpreting, wondering, and transcending one’s boundaries. It’s bound up in values, not merely in calculations. If anything, accurate intelligence is the art of giving form and substance to life’s raw material—reshaping existence through thought, invention, or art.

The Latin intelligentia itself is derived from the verb intelligere, which is composed of two parts: inter- (meaning “between”) and legere (meaning “to choose” or “to pick out”).

At its core, intelligence originally referred to the ability to “choose between” things—to discern and select among possibilities with awareness. This concept implies cognitive ability and an evaluative process, a kind of mental filtering or discernment. Intelligence, then, was tied to choice, judgment, and the act of distinguishing or “reading” reality in a meaningful way.

Greek Influence and the Notion of “Nous”

Going further back, the Latin conception of intelligence was influenced by Greek philosophy, especially by the concept of nous. Nous can be translated as “mind” or “intellect,” but it carries a more profound sense of “intuitive understanding” or “awareness.” For thinkers like Plato and Aristotle, nous was a higher form of knowing—an innate, almost divine insight that allowed one to perceive ultimate truths beyond ordinary reasoning. This notion of intelligence as an elevated, intuitive grasp of reality seeped into later Latin and Christian conceptions, making intelligence something noble and linked to wisdom.

Soul and Spirit: Medieval Christian Philosophy – A break

With the rise of medieval Christian thought, intelligence was increasingly associated with the soul and the divine. The Scholastics, notably Thomas Aquinas, took intelligentia as a faculty of the soul, distinct from mere knowledge. Aquinas emphasized that accurate intelligence came from understanding the world and divine principles, elevating intelligence as a bridge between human reasoning and God’s wisdom. This moral dimension to intelligence reinforced it as something pure and transcendent, making it a skill and a moral or spiritual calling.

The Enlightenment and the Rise of Rational Intelligence

Intelligence shed its apparent mystical and moral connotations during the Enlightenment, taking on a more secular and rational tone. As science, mathematics, and empiricism dominated intellectual life, intelligence was recast as an individual’s capacity for logical thought, reasoning, and scientific understanding. However, rational intelligence now carries the divine, with the self as the central authority instead of the church.

I think therefore I am

– Rene Decartes

What Descartes points out here is not a causal relationship; he justifies “being” as “intelligence.” That is, my intelligence is the ontological necessity of my being. Here, we have a subversion of the Christian soul and divine mechanism as a primary project for enlightenment. However, there is nothing rational about the statement. Will does not cause events to occur; it only accompanies them. As Nietzsche points out in his works, Descartes mistakenly assumes that thinking originates from a unified “I” rather than emerging spontaneously from a conflict of unconscious forces within the psyche. To say “I think” is misleading because it implies ownership and control over thoughts, whereas Nietzsche believes that thoughts arise from a multiplicity of forces within us. This understanding of multiplicity is essential for understanding why the term Artificial Intelligence is misrepresented.

In The Will to Power, Nietzsche writes: “There are no enduring substances; matter is as much an error as the ‘soul.’” He would extend this critique to the “I” in Descartes’ formulation, suggesting that it is an invention that conceals human consciousness’s chaotic and multifaceted nature.

So, the Enlightenment project now carried a falsified, measurable notion of intelligence as a natural trait, a gift from the divine self. This period also marked the beginnings of intelligence testing, where intelligence was further distilled into something quantifiable, paving the way for the modern obsession with IQ and standardized measures of cognitive ability.

Modern Psychology and the Fragmentation of Intelligence

In the 20th century, psychology further complicated the picture by breaking intelligence into emotional, social, and multiple intelligences. Intelligence was no longer a monolithic capacity but a multifaceted set of skills, each applicable in different contexts. This fragmentation reflects our modern understanding that intelligence is not merely the ability to solve abstract problems but encompasses creativity, emotional depth, and adaptability. It also mirrors a growing dissatisfaction with narrow, rationalistic conceptions of intelligence.

Machines That Learn: The New Frontier

With the development of Machines that Learn, we are witnessing the latest evolution in the concept. These bodies without organs challenge the traditional boundaries of intelligence altogether, forcing us to ask whether intelligence must be tied to consciousness, emotion, or the human experience or to be even more radical. Should we regress intelligence back to our Greek roots of intelligentia? And de-conflate the word from its Christian convolutions?

Artificial Intelligence is not only a false description of the machines that learn; it also affirms the fallacy of placing humans at the center of the world. It singularizes these multiplicities into our self-referential idea of intelligence, finding its roots at the heart of a monotheistic religion. I believe “Machines that Learn” is a more accurate description, highlighting the becoming of these multiplicities.

Furthermore, it addresses the concern of conservative scientists, who criticize that we do not know how “AI” works and that it is dangerous. I will conquer that not only do we not know how machines work, but we also do not know how the “human project” works. Both are not reasons to stop developing/enabling their becoming. The move here is towards a new mode of being/becoming. Machines that Learn no longer abide by the contemporary positivist/constructionist theology of the modern mind.

* German philosopher Friedrich Nietzsche (and after him, French philosopher Gilles Deleuze) used these categories to describe two different ways of acting and—by extension—being in the world. An action (a thought, feeling, or practice) is active when it takes something as its object; conversely, it becomes reactive when it is made the object of someone or something else. Thus, if we feel sad (or happy) and do not know why we think this way, we are reactive; if we discover why we feel this way, we can convert reactive forces into active forces. Reactive is not the same as unfavorable and should not be thought of as intrinsically evil; it is, instead, the usual state of things. It is, however, a limiting state of things because it separates us from what we can do—if we are sad for no apparent reason. If we do not seek out the cause, we are prevented from forming an appropriate response to that cause, and our power to act is reduced. We react when we could be acting, and more problematically, we still use our reaction to excuse our lack of action. Nietzsche generally refers to this state as ressentiment. Therefore, according to Nietzsche, the challenge for philosophy and life is to overcome the reactive state of things and become active, thereby constantly enhancing our power to act.


One response to “Machines that Learn: Introduction”

  1. […] discussed a position on the definition of machines that learn, what follows is a comparative study of biological neurons and how they are modeled in machines that […]