This commentary by Dileep George, from Vicarious AI, provides a critical and complementary perspective on Lake et al.'s article "Building machines that learn and think like people." George emphasizes the importance of drawing insights from human intelligence, particularly in the context of building Artificial General Intelligence (AGI). He highlights the need to understand the biological and neurological foundations of human learning and reasoning, rather than solely relying on optimization algorithms. George argues that biological evolution has developed inductive biases that are crucial for efficient and versatile learning, and that these biases can be leveraged in AI systems. He also discusses the limitations of using "human-level performance" as a benchmark for AGI, suggesting that more robust and generalizable models are needed. Additionally, George advocates for the use of message-passing (MP) algorithms in probabilistic models, which can offer a balance between speed and flexibility, similar to the inferences made in the human brain.This commentary by Dileep George, from Vicarious AI, provides a critical and complementary perspective on Lake et al.'s article "Building machines that learn and think like people." George emphasizes the importance of drawing insights from human intelligence, particularly in the context of building Artificial General Intelligence (AGI). He highlights the need to understand the biological and neurological foundations of human learning and reasoning, rather than solely relying on optimization algorithms. George argues that biological evolution has developed inductive biases that are crucial for efficient and versatile learning, and that these biases can be leveraged in AI systems. He also discusses the limitations of using "human-level performance" as a benchmark for AGI, suggesting that more robust and generalizable models are needed. Additionally, George advocates for the use of message-passing (MP) algorithms in probabilistic models, which can offer a balance between speed and flexibility, similar to the inferences made in the human brain.