❓ Can AI Become Conscious?

Consciousness remains one of the most complex and not fully understood topics in science and philosophy. It involves the ability of a being to be self-aware and to understand the world around it, including your thoughts and feelings.

For over a century and a half, scientists have been delving into the nature of consciousness, but they still haven't found a consensus on its true nature and emergence.

Sam Altman predicts that by 2030, we might see the advent of General Artificial Intelligence (AGI), capable of performing any task a human can. But the question arises: Will AGI possess consciousness? What exactly is consciousness?

Surprisingly, this question, which seems philosophical mainly, has gained significant traction over the past year. This is because, soon, we may find ourselves coexisting with machines that closely resemble humans.

📖 In his recently published book, “Putting Ourselves Back in the Equation: Why Physicists Are Studying Human Consciousness and AI to Unravel the Mysteries of the Universe,” famous science writer George Musser delves into the nature of consciousness and attempts to address whether it is an attribute solely of humans.

Musser interviewed dozens of experts, ranging from AI researchers and developers to quantum physicists, neuroscientists, and philosophers. This includes machine learning pioneer Yann LeCun and Italian physicist Carlo Rovelli, known for his work on black holes and gravity.

He presents three hypotheses on the origin of consciousness:

📌 Integrative Information Theory: This theory posits that consciousness arises from a network where all parts are tightly interconnected and measurable on a mathematical scale. For computers to achieve consciousness, their information networks must be organized similarly to neural networks in biology, utilizing feedback mechanisms.

📌 Predictive Coding: This hypothesis suggests that consciousness evolved as a survival mechanism, enabling living beings to predict events and react accordingly, such as anticipating a predator's attack. For machines to develop consciousness, they must possess similar predictive models of their environments.

📌 Neural Networks and Quantum Mechanics: Inspired by simplified models of animal brains, neural networks consist of numerous interconnected "neurons" or nodes. In these networks, especially large ones, "deep learning" processes complex information in their inner layers. The hypothesis proposes that deep learning in complex neural networks might mirror brain processes and potentially lead to consciousness.

For a deeper exploration of the concept of consciousness, we highly recommend George Musser's book, along with:

▶️ His podcast titled "Can AI Become Conscious?"

📹 YouTube interview

#mustread @hiaimediaen