In “Blake Richards’ Compelling Ideas On Intelligence, Evolution, Work And Misplaced Fears About AI,” the author explores the multifaceted perspectives of Blake Richards, an Associate Professor at McGill University and a renowned expert in AI and brain neurology. The article, based on an interview with Jon Krohn for the Super Data Science Channel, delves into Richards’ views on intelligence in both humans and machines, the evolution of work, and the misconceptions about AI’s dangers. Richards, differing from his mentor Geoffrey Hinton, believes that the fears about AI advancing too rapidly are misplaced and that the real danger lies in “stupid AIs.” He emphasizes the importance of understanding intelligence as an adherence to norms, extending beyond cognitive abilities to include societal and survival behaviors. Richards also expresses skepticism about the concept of Artificial General Intelligence (AGI) as a singular capability, instead advocating for a multifaceted approach to evaluating AI’s intelligence across various metrics.
The article also discusses Richards’ stance on biomimicry in AI development, where he argues for functional mimicry over biological mimicry. This approach focuses on replicating brain functionalities, such as episodic memory, rather than biological processes. Richards’ insights into AI’s potential evolution and his optimism for AI systems improving across a broad range of tasks are highlighted. Additionally, the article touches on Richards’ views on AI’s existential risks, where he downplays fears of AI dominance, suggesting that AI’s evolution will likely be guided by cooperation and mutualism rather than competition. He also advocates for practical safety measures in AI development, emphasizing the need for independent auditing over restrictive governmental regulations. Richards foresees a future where AI augments human capabilities, particularly in creative and intellectual tasks, rather than replacing humans, suggesting a symbiotic relationship between AI and humanity.