Has AI Hit a Dead End?

Abstract

95 percent of brain activity has been found to be fluctuations seemingly unrelated to conscious activities involving sensory perception, motor actions and cognition. In the neuroscience framework they are interpreted as noise. Since fluctuations are poison for deterministic computation, the finding poses a serious problem for model of the brain as a deterministic classical computer. In this article the TGD based interpretation of the long range fluctuations as quantum fluctuations characterized by the value of the effective Planck constant h_eff = nh₀ labeling the phases of ordinary matter identified as dark matter and residing at magnetic body (MB) of the system is discussed. n has number theoretic interpretation and can be regarded as a universal IQ so that fluctuations are a prerequisite for intelligence. According to the TGD based view about neuroscience primary sensory percepts reside at the sensory organs which requires back and forth communications between brain and sensory organs to build sensory perceptions as standardized mental images. These communications must be fast and the proposal is that they use dark photon signals. In this view nerve pulses do not represent signals inside the brain but act as neural relays at synaptic junctions making possible long range dark photon communications inside the brain. Part of the metabolic energy associated with the fluctuations could be used to the building of mental images in the proposed manner. Nerve pulse patterns generate Josephson radiation communicating sensory information to MB and also require metabolic energy. Dark cyclotron radiation from MB represents control signals to the brain. In both cases, long range fluctuations at brain level are involved.