Artificial Intelligence and the Brain: Two Ways of Thinking We live in an era when machines are beginning to think. No, they do not feel or experience, but they can write poems, diagnose diseases, drive cars, and even conduct dialogues that are almost indistinguishable from human ones. Artificial intelligence has burst into our lives and made us ponder: what, after all, makes us human? What is the difference between our brain and a neural network? And is there anything in common between them besides the word “neuro”? World Brain Day is the perfect occasion to delve deeper into this and try to understand where biology ends and code begins. Architecture: biological chaos versus mathematical order The first and main difference is how both “processors” are structured. The human brain is the result of millions of years of evolution. It is not designed but grows like a living organism. Its neural networks are not perfect: they are noisy, slow, prone to fatigue, injury, and aging. But it is this imperfection that makes it flexible. The brain can learn from a single example, it is capable of generalization, it knows how to transfer skills from one area to another. It is a living system that constantly reconfigures itself under the influence of experience. Artificial intelligence, on the other hand, is created by engineers. Its neural networks are mathematical models operating on digital carriers. They are accurate, fast, and predictable. They do not get tired or sick. But they cannot go beyond the data on which they have been trained. They do not understand context unless it has been embedded in the training. Their “flexibility” is just the ability to iterate through billions of combinations but not to create new principles of thinking. This comparison is reminiscent of the difference between a living tree and its 3D model. The model is beautiful and accurate, but it does not grow and bear fruit. The tree is chaotic, unpredictable, but it is alive. Learning: experience vers ... Read more