Re:What Happens in a Mind That Can't 'See' Mental Images
by ninetyninenine on Friday August 02, 2024 @06:49PM
(#41143333)
We have algorithms that can hallucinate now.
LLMs basically are examples of how humans visualize things. With a few differences. Humans have more fine grained control over the result and understanding of a query. LLMs have greater detail in the sense that the LLMs knows the location of every wrinkle on a face while our imagination delivers an approximation with detail only being rendered if we decide to focus on the details.
LLMs basically are examples of how humans visualize things. With a few differences. Humans have more fine grained control over the result and understanding of a query. LLMs have greater detail in the sense that the LLMs knows the location of every wrinkle on a face while our imagination delivers an approximation with detail only being rendered if we decide to focus on the details.