Stupid Artificial Intelligence?
Millions of words have already been written about the digital technological revolution and the phenomenon of Artificial Intelligence. Many more will be written, including those on the growing development of so-called LLMs, large language models used by Generative Artificial Intelligence (AGI), trained on huge amounts of textual data to understand, generate, and analyze language in a similar way to a human being. The impact on our lives, on the economy, on anthropology itself, on our view of man, his nature and his role in the world, is immense and growing day by day.
Since this epochal revolution affects every aspect of existence, to the point of changing the meaning of terms such as brain, intelligence, and consciousness, and irreversibly changing the map of the economy, power, and the human condition itself, every branch of human knowledge looks at the phenomenon of Artificial Intelligence from its own specific point of view. It is not a question of slowing down its progress—that would be like trying to block the flow of a river with one hand—but of understanding and evaluating it as a tool, not as an end in itself or a formidable mechanism of power.
Let’s start with the definition: Artificial Intelligence is the ability of a machine to exhibit human faculties such as reasoning, learning, planning, processing, and problem solving autonomously. AI allows machines to analyze data to make decisions, make predictions, act, and interact with the environment, automating “human” processes. The danger that looms over us concerns the possibility that machines will not only replace human functions, but will end up weakening, reducing, and causing many of our abilities to disappear. Some point out that AI does not imitate all the faculties of our brain, but only those of one of the two hemispheres of which it is composed, the left. Technologists and computer scientists have studied the neurological structure and functioning of the human brain with great attention. AI becomes the omnipotent technological interface of contemporary dominant ideas, whose preference for analytical, instrumental, mechanical, and taxonomic thinking is evident. From a neurological point of view, this is the victory of the left brain. This is the thesis—obviously unrelated to AI—of Scottish philosopher, psychiatrist, and neuroscientist Iain McGilchrist, expressed in his monumental treatise The Master and His Emissary (2009), whose subtitle is The Two Halves of the Brain and the Making of the Western World. An influential economic journalist, Greg Ip, a leading writer for the Wall Street Journal, takes up this thesis and relates it to AI. According to McGilchrist, Western man is dominated by the version of the world created by the left hemisphere and has forgotten the insights produced by the right.
We need both hemispheres, but above all, McGilchrist argues, the left hemisphere (the emissary) must work in the service of the right (the master). Each cerebral hemisphere deciphers reality in a coherent way, but incompatible with the other. The right hemisphere experiences the world in its entirety, complexity, and organicity, leaving out the details, while the left hemisphere is more analytical and therefore fragmentary. The signs of this internal conflict are visible in the history of our civilization. In the disembodied world dominated by digital technologies, the left hemisphere is dangerously gaining the upper hand over the right, perhaps changing the way we think and understand reality forever. The West has experienced an oscillation between the two modes: certain periods have been influenced by the right hemisphere, for example the Renaissance and Romanticism, others by the left, such as the Counter-Reformation or the Enlightenment. The leap from biology to culture is based on two arguments: mimesis and epigenetics. The mimetic attitude can be observed from an early age: children imitate the behavior of adults; imitation of a model develops individuality, as well as producing enculturation. The dominant hermeneutics in contemporary times records the triumph of the left hemisphere over the right, which abstracts meanings from context and advances neurotically with irrepressible optimism, typical of the left hemisphere. It reasons in pairs of opposites: true, false, black, white, and so on.
The right brain is aware of complexity, nuances, and the coexistence of opposites; it prefers to see the whole organically rather than dissect it. In computer language, it is continuous as opposed to discrete. The left hemisphere is a brilliant emissary, capable of logically reconstructing, step by step, the intuitive leaps of the right brain, but it cannot replace it. Artificial intelligence has taken on the cognitive mode of the left hemisphere alone, mutilating half of the model of natural human intelligence. The risk is simplification, the reduction of everything to segments, infinite sequences of zero and one, open and closed, structurally incapable of grasping the whole, the complexity, the infinite range of colors of the human palette. As the artificial model is overlapping with the original to the point of replacing it, the risk is an epigenetic mutation that restricts the human by weakening the functions of the right hemisphere.
The outcome could be a humanity tailored to Artificial Intelligence. The world turned upside down, even in the field of technology. AI will replace human intelligence, giving way to the tyranny of machines. The economy is heading towards an unprecedented replacement of human labor, leading to mass unemployment and serious social conflicts. These are the valid objections that mark the advent of AI. A further consequence is that the prevalence of the artificial model will lead us to intellectual stagnation. Language models such as ChatGPT, Google Gemini, and Anthropic’s Claude are very effective at locating and connecting knowledge. But they add nothing to the body of knowledge. In other words, they are not curious. They lack a sense of wonder, amazement, and human abstraction. For this reason, the architecture of AI may not be prepared to discover new knowledge. To use McGilchrist’s lexicon, AI would be the left side of the human brain transferred to a machine. Calculating, analytical, and reductive, it treats the world as a giant data warehouse. Processing can provide us with a powerful technological improvement over what already exists. But left-brain reasoning is incapable of analogy, imagination, or emotional thinking. Perhaps ‘true’ artificial intelligence would need a mood button capable of switching from a state of hyperactive concentration to one of meditation.
It is unlikely that a switch could activate a state of concentration on a seemingly irrelevant issue in AI, a mental state that often triggers the creative process. Geneticist James Watson explained that he intuited the double helix structure of DNA while playing tennis. We should consider AI a valuable aid that allows the human user to focus on issues that require imagination, logical and cognitive leaps of the mind, or act as an emissary of the master, the right hemisphere of the brain with its capacity for intuition, creativity, and analogy. Relying on AI atrophies critical thinking just as relying on GPS weakens spatial memory. Statistical evidence shows that cognitive functions in humans decline in direct proportion to reliance on AI and Internet search capabilities.
Teachers know this from daily experience: they observe declining attention spans, weakened memory, a crisis in critical thinking, and a lack of curiosity. AI will bring progress in some areas, but in the long term, the result could be scientific stagnation: we will see a refinement of existing knowledge and an accelerated exploitation of its potential, but there may be less and less new knowledge. Artificial intelligence could also accelerate the decline of truth that we are already experiencing with dismay. A Wikipedia editor has described in detail how he deliberately introduced biased information into the digital encyclopedia, the bible of contemporary life. LLMs encourage similar operations on a much larger scale. One example is Google Gemini’s report on Matthew Shepard. “He was a gay college student who was brutally beaten and left for dead in a hate crime in Laramie, Wyoming, in 1998”. The investigation showed that his murder was not motivated by hatred of his sexual orientation. But this is how pro-gay propaganda presented Shepard’s murder for years. Given the enormous amount of text published by activists and those misled by propaganda, LLM models of Artificial Intelligence will tend to repeat this widely disseminated but false narrative.
Blind faith in AI makes its statements incontrovertible truths: technology says so! Organized movements, governments, and other centers of power are working to put ideologically distorted or false material online in order to shape the processing of LLMs, which will then present the narrative as established fact. These maneuvers determine and will increasingly determine the contamination of content, as enormous amounts of AI-generated text—designed to influence results—will flood the web. The result will be a degradation of knowledge that will dwarf the damage caused by television and social media. The transformative potential of AI has powerful effects, but change is not always synonymous with progress. Especially if it reproduces – extending it indefinitely – not the whole of human intelligence, but only the instrumental part, located in the left hemisphere of the brain, the only part that interests its masters. Stupid Artificial Intelligence devoid of meditative thought.