by José M. Muñoz , Javier Bernacer, Alva Noë, Evan Thompson • 11 days ago
The article discusses the limitations of large language models (LLMs) in achieving human-level intelligence, specifically highlighting three key abilities they may lack: generalization, representation, and selection. These shortcomings suggest that artificial general intelligence (AGI) may remain unattainable for LLMs. The authors argue that these cognitive functions are essential for problem-solving and decision-making, which are critical components of human intelligence.