Abstract
Knowledge representation formalisms aim to represent general conceptual information and are typically used in the construction of the knowledge base of reasoning agents. A knowledge base can be thought of as representing the beliefs of such an agent. Like a child, a strong-AI (AGI) robot would have to learn through input and experiences, constantly progressing and advancing its abilities over time.
To emulate human intelligence, we need not only the statistical AI generated by neural networks but also the concept of causality of events translated into directionality of logic entailments and deductions. Moreover, by using axioms, we can guarantee the controlled security of the robot's actions based on logical inferences.
For AGI robots, we consider the 4-valued Belnap's bilattice of truth-values with knowledge ordering, where the value 'unknown' is the bottom value. Sentences with this value represent unknown facts, which are the missing knowledge in AGI robots. Thus, these unknown facts are not part of the robot's knowledge database, and through learning from input and experiences, the robot's knowledge would naturally expand over time.
Consequently, this phenomenon can be represented by the Closed Knowledge Assumption and Logic Inference provided by this paper. Additionally, the truth-value 'inconsistent', which is the top value in the knowledge ordering of Belnap's bilattice, is necessary for strong-AI robots to support inconsistent information and paradoxes, such as the Liar paradox, during deduction processes.
Blogger's Review: This paper presents a novel perspective on the closed knowledge assumption, enhancing the reasoning capabilities of strong AI robots. It emphasizes the importance of causality in the learning process, providing a theoretical foundation for addressing complex logical paradoxes, showcasing the potential of neuro-symbolic approaches.