Reflections on Values, Meaning, Logic, and Technology
Artificial Intelligence (AI) is a raw analytical machine — a technology capable of performing analytical operations with immense speed and precision. However, artificial intelligence lacks the capacity to make qualitative value judgments or assessments or to understand meaning. These capabilities, I would argue, are the fundamental features of consciousness and awareness.
Logic, on the other hand, is a philosophical discipline that seeks to discover truth through analytical reasoning. I believe it represents the intersection between human consciousness and analytical operations.
The word “logic” originates from the Greek word logos, which has a variety of translations, such as reason, discourse, or language. Logic is traditionally defined as the study of the laws of thought or correct reasoning, and is usually understood in terms of inferences or arguments. Reasoning is the activity of drawing inferences. Arguments are the outward expression of inferences. An argument is a set of premises together with a conclusion. Logic is interested in whether arguments are correct, i.e. whether their premises support the conclusion.
Wikipedia entry on Logic – Definition
AI and the Tools of Applied Logic
An essential feature of the logical methodology is the truth table, which defines how analytical operations can be applied to qualitative value judgments to extract meaningful inferences. The logical operations described by truth tables, such as “and”, “or”, and “not”, translate into binary operations in the world of computer analytics. These same operators apply to both the binary digits or bits that constitute computer circuitry at its most basic level, as well as to the qualitative values of T and F, as expressed by logical operations in truth tables.
However, while computers, including AI, can perform such analytical operations with tremendous speed and power, they cannot fundamentally distinguish between truth and falsehood or comprehend the meaning of these qualities. Computers and AI can perform deductive, inductive, and other analytical processes, aided by programs, algorithms, and logic circuits, to deliver the inferences we need. However, computers, including AI, must base their operations on premises and presuppositions that they cannot independently assess or verify; they must be instructed beforehand whether these premises are intrinsically true or false.
In the world of computers, this concept is known as GIGO, an acronym for “garbage in, garbage out.” In other words, the quality of the results you get from computers, including the most advanced generations of AI, fundamentally depends on the quality of the data and premises you feed into computer algorithms and programs. Computers and AI lack the independent ability to make the value judgments required to assess the data and premises on which they operate. They blindly apply analytical processes and operations to the premises provided, whether sound or faulty, arriving at results that may correspondingly be sound or faulty.
Human consciousness, on the other hand, is decidedly different. While far from perfect, conscious human beings are capable of making value judgments and understanding the meaning of things. Human beings can, to some degree at least, inherently, perhaps innately, discern between truth and falsehood, right and wrong, good and bad, etc. Perhaps this ability stems from the built-in human moral compass, or it may be the result of human intuition. Regardless, human conscious awareness appears to have the mysterious, even mystical, ability to make value judgments and to understand or grasp meaning. It may not always be accurate, but it is a human faculty nonetheless — a feature of human consciousness.
AI, even a so-called “superintelligent AGI (i.e., artificial general intelligence),” will never possess the equivalent of this human ability. It may well develop highly sophisticated analytical or logical circuitry, with the ability to perform calculations, in effect, instantaneously and with perfect accuracy — but it will never be anything more than that.
If we understand logic to be the amalgam of meaning and analysis, it follows that AI is incapable of doing logic.
AI can perform analytical operations (e.g., “and”, “or”, “not”, “if-else-then”) with unimaginable speed and accuracy. Still, it will never be able to derive value or meaning from the results of its operations or, indeed, to verify their validity or accuracy. AI may be able to perform raw logical and mathematical computations, assuming that its algorithms and programming are sound and accurate. However, AI can never understand what logic means.
Analytical vs. Holistic Thinking
The built-in human faculty for conscious value judgments, truth detection, and the discernment of meaning is complex and mysterious. It may be partly connected to sensory perception, partly tied to one’s moral compass, partly associated with intuition, perhaps linked to the pineal gland (which some posit as the seat of human intuition), and partly learned or developed through education, experience, and meditative practices. This is why it is not necessarily easy to misguide, misinform, brainwash, or indoctrinate even the most gullible human subjects. Even a child at a relatively early age possesses a degree of innate skepticism and incredulity.
Computers, including the most advanced AI machines, have no such faculty. This is precisely why we see embarrassing AI glitches in even the most advanced AI learning models. The reason for this is the basis of the GIGO principle in computer programming — computers, including AI, have no inherent faculty or ability to distinguish between truth and falsehood, between the logical qualities of T and F. They rely entirely on the premises fed to their training models. If an AI is trained on flawed, faulty, or corrupted premises, even a highly advanced AI will glitch out and arrive at flawed conclusions. AI lacks a built-in mechanism to signal to itself that it has failed, as it operates purely analytically.
Human beings possess the capacity for holistic, aesthetic, right-brained thinking, which can question or conflict with left-brained, linear, analytical thinking. AI, on the other hand, is constituted purely of linear, left-brained, analytical operations with no holistic, right-brained capabilities whatsoever.
It would appear, therefore, that there is a mysterious dimension to human consciousness beyond purely analytical thinking. There are the intuition, the subconscious, the collective unconscious (as identified by psychologist Carl Jung), dream states, the capacities for aesthetic appreciation and value judgment, the moral compass, and so on. Human beings have a sensory apparatus and an endocrine system (especially the pineal gland), which also influence the mysterious and complex faculties of human conscious awareness.
AI, on the other hand, is a pure analytical machine. It may be extremely powerful in what it can do, but it has no mysterious abilities beyond its programming, which it is incapable of challenging or questioning. AI is a logic circuit without the value judgments or the comprehension of meaning that form the premise and basis of logic. As such, AI relies on human input to provide the value judgments it requires to deliver accurate and meaningful results.
AI and the Mechanics of Reading
When human beings read a book, they are engaging in something akin to a mystical process. Reading implies deriving and understanding meaning from every word in a text, as well as the meaning associated with the broader concepts behind each phrase, each sentence, each paragraph, each chapter, and the entire book as a whole.
When AI consumes and analyzes a text, on the other hand, it has no comprehension of meaning or value at all. It performs analytical operations and statistical computations to parse the text and its language, arriving at its conclusions at what may be lightning speed. However, the AI lacks a genuine understanding of the text underlying this superficial analytical level. It has no deep “human” understanding of the text.
When we use AI to assist us in our creative pursuits, such as to provide us with ideas for writing, or to summarize the contents of a text, or to research a subject, what we are doing, in effect, is surrendering our own abilities to assess values and comprehend meaning to the values and understanding of the AI’s trainers. We are relying on the perceptions and prejudices of the programmers who may have trained the AI we are using. These individuals may not be the most knowledgeable or perceptive, and indeed, they may not share our values. They may not have the best taste, discernment, or understanding. In fact, from our own perspective, they may be utter philistines.
Nevertheless, the AI trainers, however untutored and pedestrian they may be, have infused their own limited values, judgment, tastes, and understanding into the AI they have trained, which we are now using. These values, tastes, and so on, immeasurably amplified by the AI’s logical circuitry, are subsequently offered to us as a service, so that we no longer need to rely on our own judgment, understanding, abilities, learning, humor, depth perception, and so on. We are instead given the option to divest ourselves of the effort and responsibility of thinking for ourselves and to allow a soulless machine to do all our thinking for us.
Reading a text is a transcendent, mystical experience of communion with another conscious mind — the mind behind the creation of the text we are reading. The human mind is capable of grasping or comprehending the meaning of a text at various levels — the denotative or face value sense of the text, its connotative subtext, its figurative, symbolic or allegorical meanings, its allusions, its hidden meanings, its double entendres, its “higher meaning,” its ironic sense, its broader contextual meaning, its holistic, thematic meaning, and so on.
AI, on the other hand, can perform the mechanical operations of parsing and analyzing a text with tremendous efficacy — but does it have even a basic comprehension of the face-value meaning of the words in the text? I don’t believe it does — AI has no ability to comprehend the meaning of the words or language in the text at even the most basic, fundamental levels, in any real sense, because AI lacks the capacity for comprehending any meaning or making any value judgments.
Preserving Humanity
As such, if we relinquish our humanity and human abilities to AI, including our skills to think, read, learn, comprehend meaning, and make value judgments for ourselves — if we allow these abilities to atrophy and disappear — we may well lose them from our experience entirely. AI does not inherently possess any of these abilities. If they are missing in the AI’s human trainers and programmers, these capabilities may disappear entirely from human discourse and experience over time. This is why they must be cultivated and inculcated in human beings from an early age and protected from the soulless, inhuman influence of AI.
