Why Computers Can’t Be Conscious.

🌈 Abstract

The article discusses the limitations of neuroscience and computer science in understanding human consciousness and the mind. It argues that the brain does not generate consciousness, and that the mind is more than just a physical system. The article also critiques the reductionist and materialist approaches of modern neuroscience and AI, and suggests that consciousness and meaning are fundamentally different from the physical world.

🙋 Q&A

[01] The limits of science and science

1. What are the key issues with the scientific approach to understanding consciousness?

• Neuroscience cannot answer the "hard problem" of consciousness - how matter can give rise to subjective experience.
• Science can only describe and predict how nature behaves, not what nature is. Understanding the meaning and interpretation of consciousness belongs to the realm of philosophy and metaphysics.
• Quantum mechanics shows that science deals with the empirical realm of reality, while the meaning of quantum reality is a philosophical question.
• Consciousness is a subjective and non-reducible phenomenon that cannot be fully explained by its objective correlates.

2. What are the limitations of the scientific approach to understanding the mind?

• Science reduces all reality to quanta, numbers, or units of measure, and deals with the exterior, empirical realm of reality.
• To understand the meaning of a text like King Lear, we need interpretation and intersubjective understanding, not just empirical explanation.
• There are no instruments to directly observe reason, values, thoughts, or emotions. Understanding the mind requires dialogue and apprehending meaning and context.
• The belief that matter generates qualia (subjective experience) is a philosophical doctrine, not a scientific fact.

[02] Is the mind a machine?

1. What are the issues with the view of the mind as a computational machine?

• Biologism and neurophilosophy conflate correlation with causation and identity, and put mind, biology, and matter at the same abstraction level.
• The presence of neural correlates of subjective experience does not establish that the brain generates consciousness.
• Computers can manipulate symbols according to rules without actually understanding meaning, as demonstrated by the Chinese room thought experiment.
• Syntax (structure, form, rules, and arrangements of words) is not equal to semantics (meaning and interpretation).

2. How do the concepts of information and memory challenge the computational view of the mind?

• The engineering definition of information, which excludes meaning, enables panpsychist philosophers to see information and experience everywhere, even in rocks and subatomic particles.
• Neuroscience has not found the location or "address" of memory "files" in the brain, and the theories of memory are still divided between local and distributed models.
• Connectionism, while offering an alternative to the cognitivist view of the mind, still does not explain how the emergent properties of neural networks relate to symbolic computation or how they give rise to consciousness.

[03] Fallacious abstractions and misplaced isomorphisms

1. What are the issues with the appeal to emergence, patterns, and complexity in explaining consciousness?

• Appeals to emergence, patterns, and complexity have enormous fallacious explanatory power, as they do not explain the "jump" in qualitative essence from matter to life to mind.
• The mind finds and defines patterns, emergence, and complexity, so appealing to these concepts puts the cart before the horse.
• The complexity of neural patterns in the brain should not be surprising that consciousness "emerges" from it, as the criteria for complexity, pattern, and emergence are themselves not understood.

2. How do the phenomenological and existentialist perspectives challenge the computational view of the mind?

• We are not a brain in a vat, as the embodied cognition movement and AI research implicitly conceive of intelligence and consciousness.
• Physical interaction with the environment is essential for proper neural development and perception, as demonstrated by the example of the visually impaired kitten.
• Our consciousness and self are not stored inside a body or cabinet, but are part of a meaningful life-world that is co-created with others.
• Reasoning from a computational perspective is significantly simpler than computing perception and sensorimotor skills, as demonstrated by the challenges in building autonomous vehicles.

[04] Conclusion

1. What is the key argument made in the conclusion?

• Consciousness and mind are much more than a mirror of nature, as cognition is the enactment of a world and a mind based on a history of actions.
• The challenge posed by phenomenologists remains - the brain is perceived, dissected, and experimented on by and through the minds of scientists, not the other way around.
• The belief that science can create consciousness is a "Disneyfication of Artificial Intelligence", as the "ghost in the machine is just that: a ghost."
• The hard problem may be not how consciousness arises from matter, but how matter arises from consciousness, as we only know of things that exist within our experience.