"AI should NOT be regulated at all!" - Prof. Pedro Domingos
Machine Learning Street Talk (MLST)
Symbolic Learning: The Foundation of AI Insights
Symbolic AI is distinct from knowledge engineering and expert systems, encompassing a broader category called symbolic machine learning, which has roots dating back to the 1970s and earlier. Inductive logic programming stands as a crucial aspect of this field, showcasing the evolution of symbolic approaches. Key components, such as inverse deduction, trace back to historical figures like Jevons, revealing the long-standing relevance of these concepts. Within the machine learning community, conferences like NeurIPS and ICML reflect the divide between connectionist and symbolic paradigms. All learning, regardless of method, requires prior knowledge, as highlighted by the no free lunch theorem, emphasizing that success hinges on the type and amount of prior knowledge incorporated into algorithms. Different tribes in AI, such as Bayesians, connectionists, and symbolists, vary based on the models or knowledge bases they utilize, with symbolists often being more explicit about their biases and the rules governing learned outcomes.
Remember Everything You Learn from Podcasts
