Part 8/14:
However, most AI systems today—such as large language models—are simply amplifying human capabilities rather than surpassing them in an unexpected or incomprehensible way. If an AI can only do what humans can do but faster or more efficiently, concerns about superintelligence are less immediate.
The Human Mind as a Benchmark: Unpredictability and Constraints
Despite the unpredictability within humans, we generally consider humans as a stable agency because of our physical and cognitive constraints. Our limits—energy, focus, processing speed—act as natural brakes on potential destructive actions.