Your Brain Isn't Hardware Running Software

The brain's function extends beyond mere computation, challenging the metaphor of the brain as hardware running software due to the inseparable nature of its physical substrate and processes.

• The brain as a computer is a metaphor, not a literal description.

• Biological processes challenge 'substrate independence' due to the inseparability of physical form and function.

• Understanding cognition requires moving beyond strictly algorithmic and Turing-machine-centric views.

The idea of the brain as a computer is a deeply embedded but misleading metaphor, as biological systems like the brain do not neatly separate 'what they do' from 'what they are.' This challenges the concept of 'substrate independence,' which assumes mental processes can be transferred to any material. A richer understanding of both biology and machines is needed, moving beyond strictly algorithmic views.

Brain-as-Computer Metaphor

• 00:01:35 The prevailing view that the brain is a computer made of meat, with cognition and consciousness solely based on computations, leads to the assumption of substrate independence. This implies that the material composition doesn't matter, only the computations. However, this perspective overlooks that the brain-as-computer idea is a metaphor, not an accurate representation, hindering a full understanding of biological systems.

Inseparability of Brain & Mind

• 00:01:52 In biological systems like the brain, there is no clear distinction between the 'mind' and the 'wetware,' or 'what it does' and 'what it is.' This inseparability challenges the fundamental assumption that computation alone matters, as computation requires a sharp division between software and hardware. Without this clear separation, the idea that brain functions could be instantiated on a different substrate becomes less tenable, even though simulations remain useful tools.

Limitations of Computation Paradigm

• 00:03:23 The computational paradigm and algorithmic notions are formal models that do not fully capture everything needed to understand life, or even complex machines. The traditional view limits imagination about what machines could be, neglecting alternative histories of AI rooted in cybernetics, dynamical systems, and emergent properties. Even seemingly algorithmic processes like 'bubble sort' can exhibit emergent properties not fully captured by a strict algorithmic view.

Beyond Turing Computation

• 00:09:42 Biological systems, and even some machines, operate beyond the scope of a universal Turing machine. Processes that are stochastic or continuous, or functions that inherently involve the transformation of specific material substrates like metabolism, are not purely computational in the Turing sense. These 'non-Turing processes' extend beyond 'hypercomputation' (solving problems a Turing machine cannot), suggesting a broader class of operations critical for understanding complex systems.