Minding the gap between artificial and biological computing paradigms for biologically loyal AI

Area of Science:

• Neuroscience and Artificial Intelligence
• Computational Theory
• Cognitive Science

Background:

• The theoretical foundations of neuroscience and artificial intelligence (AI) are disparate, hindering progress in AI development.
• Computability theory, a cornerstone of computer science, is narrower than cognition, creating logical limitations for computational biology and AI.
• Existing computational models may limit our understanding of brain systems and the development of advanced AI.

Purpose of the Study:

• To identify and analyze mathematical and biological activities neglected by current computability theory.
• To bridge the gap between neurobiology and artificial general intelligence (AGI).
• To propose a new theoretical paradigm for integrating neuroscience and AI.

Main Methods:

• Examination of mathematical activities, such as theorem proving beyond current computational logic.
• Analysis of neuronal functions exceeding the complexity of simple transistors, informed by recent neurobiological discoveries.
• Survey of potential frameworks and inspirations for a novel synthesis of AGI and neurobiology.

Main Results:

• Identified key cognitive and biological processes not encompassed by traditional computability.
• Highlighted the complexity of neuronal functions compared to artificial computing elements.
• Established the need for a paradigm shift in AI development that incorporates neurobiological principles.

Conclusions:

• A new computing paradigm is necessary to effectively integrate neuroscience and AI.
• This paradigm must account for mathematical and biological activities beyond current computational limits.
• The hypothesis suggests that a successful AGI will require thorough integration of cognition and motion.