Combining generative AI with rules-based systems could lead to artificial general intelligence

Old AI approaches, previously dismissed as unable to achieve Artificial General Intelligence (AGI), may be resurrected. Experts suggest combining these "rules-based systems" with modern generative AI could lead to AGI. Generative AI and large language models, once hailed as the path to AGI, are now facing scrutiny for their scalability limitations. This has prompted a re-evaluation of older AI methods, like expert systems, which relied on symbolic logic. The proposed "neuro-symbolic AI" or "hybrid-AI" integrates neural networks with rules-based systems. While some fear a return to outdated methods, this approach offers a potential new direction towards achieving human-level artificial intelligence.