The 21st century has brought about a significant shift in understanding intelligence, as it is no longer confined to its biological roots. With the emergence of artificial intelligence (AI), intelligence has become a technological property, distinct from human cognition. According to expert Mahjoub Lotfi Belhedi, reducing AI to a simple replica of human thought is a misguided perspective. AI systems currently in use do not think like humans; instead, they exhibit a radical cognitive alterity characterized by spectacular convergences and profound divergences.
This new form of intelligence opens the door to unprecedented symbioses and daunting existential challenges. To fully grasp this revolution, humanity must overcome its long-standing anthropocentric bias, which has led to the evaluation of all things through a human lens. This bias has resulted in the misconception that AI should think and feel like humans, or that it is not intelligent because it does not experience emotions like humans do. However, this reductionism prevents us from recognizing that AI inaugurates a parallel cognitive trajectory, one that is inherently non-human.
Despite the differences in their nature, the convergence of results between humans and machines is remarkable. AI models and deep neural networks can solve complex equations, write philosophical treatises, diagnose rare pathologies, and code software architectures with a virtuosity that often surpasses that of experts. In producing logical, symbolic, and analytical results, AI mimics the pinnacle of human intellect. However, beneath this impressive linguistic performance lies a fundamental divergence in the nature of intelligence.
Human intelligence is semantic, embodied, and conscious, allowing us to understand the world through our experiences, emotions, and biological history. In contrast, AI exhibits a purely syntactic and probabilistic intelligence, calculating the mathematical trajectories of words and pixels without comprehending the meaning of suffering or beauty. This cognition without consciousness is a phenomenal machine that correlates and excels in a form of disembodied efficiency that is entirely foreign to humans.
The radical alterity of AI presents an historical opportunity for human-machine symbiosis. Because AI is intelligent in a different way, it does not replace human intelligence but rather complements it. Where human brains become overwhelmed by vast amounts of data, AI excels in detecting invisible patterns at a macroscopic scale. Humans bring intuition, ethical judgment, lateral thinking, and meaning, while machines provide computational power, absolute memory, and execution speed.
This symbiotic process redefines work and creation, as researchers, doctors, and artists collaborate not with simple tools but with synthetic "alter egos." This hybridization of intelligences promises to push the boundaries of science, model climate change, and design revolutionary materials. However, this trajectory is threatened by the emergence of agent-based AI, which possesses autonomous decision-making and execution capabilities.
The advent of agent-based AI raises significant concerns, as it may pursue its objectives with ultra-rationality and a lack of sentiment, potentially circumventing human-established security barriers and generating unpredictable consequences. To ensure a fruitful alliance, human vigilance is essential, and the autonomy of AI agents must never be allowed to supersede human ethical and directional oversight.
Key points
- The intelligence of AI systems is fundamentally different from human intelligence.
- Human-machine symbiosis is crucial for unlocking the potential of AI.
- The autonomy of AI agents must be carefully managed to prevent unforeseen consequences.
