Sam Altman: AI Threatens Our Cognitive Abilities

Sam Altman Warns About the Cognitive Risks of AI

Sam Altman, the CEO of OpenAI, recently expressed his concerns regarding the potential impact of artificial intelligence on our cognitive abilities. According to him, AI could lead to cognitive atrophy if individuals rely too heavily on it to perform intellectual tasks. Altman emphasizes the importance of continuing to stimulate our brains, even as AI takes on more and more functions. He envisions a future where AI is ubiquitous and personalized, but he asserts that human judgment will remain indispensable.

During an episode of the "Invest Like the Best" podcast, Altman highlighted that one of the often-overlooked dangers of AI is the possibility that people will let technology think for them, which could weaken their mental capabilities. He posed the crucial question: "How are we going to avoid cognitive atrophy?" This question arose when he was asked what the most important unresolved issues regarding AI are.

Researchers Concerned About AI's Impact on Skills

Altman went on to wonder how we could use these tools while continuing to exercise our brains and understand what is truly important. His comments come in a context where many researchers warn that while AI may enhance worker efficiency, it could also undermine their underlying professional and cognitive skills.

Nataliya Kosmyna, a researcher at MIT, shared her perspective by stating that no previous tool has offered such an opportunity to delegate thinking. In a preliminary study, she found that individuals using generative AI to write essays performed worse over time compared to those using Google or no assistance at all.

Cognitive Substitution: A Risk for Cognitive Reserve

Vivienne Ming, a theoretical neuroscientist and AI researcher, described the phenomenon where people abandon cognitive work in favor of AI as "chronic substitution." She explained that this repeated substitution could weaken cognitive reserve, which is a key defense against dementia. These concerns are becoming increasingly pressing as AI tools evolve to perform more complex tasks, ranging from document review to analysis and customer support.

A Promising Future but with Challenges

Sam Altman is not the only one expressing concerns. Arthur Mensch, CEO of Mistral AI, also warned about the "deskilling" of workers, meaning the gradual loss of skills due to increased reliance on technology for information retrieval.

However, Altman also shared an optimistic vision of the future during his interview. He imagines a "pretty magical" world where a personal AI could be constantly active, capable of reading documents, listening to meetings, remembering old emails, and using available computing power to generate ideas and suggest future actions.

Nevertheless, he clarified that even the most advanced AI models may struggle to replicate distinctive human qualities. "The world may need a new term to describe the type of judgment in which humans excel, and that AI seems to have difficulty mastering," he concluded.