“Artificial Superintelligence (ASI) is a hypothetical stage of AI where a computer system surpasses the cognitive performance and problem-solving abilities of the most gifted human minds across every domain of interest.”
that’s an “industry term” in the same way that “panopticon” is a prison industry term, or that “brain in a vat” is a neurology term. It’s a name made up by science fiction writers and philosophers to describe a thing that does not exist, and has never existed. Might it exist in the future? Maybe, but that’s irrelevant to an issue of law. You can’t pass a law banning an ill-defined “panopticon”, because that isn’t a term of art enshrined in any US code.
because, as you say, it is purely hypothetical, such a law would be unenforceable, and would likely fail a test in court.
because it is, once again, nonexistent as of yet, the burden is on you to prove that they aren’t specifically referring to the term which the Shit lord of Hamberder enshrined in his executive declaration, because that use of “Superintelligence” actually has a basis in current US government parlance.
It’s a real term.
“Artificial Superintelligence (ASI) is a hypothetical stage of AI where a computer system surpasses the cognitive performance and problem-solving abilities of the most gifted human minds across every domain of interest.”
I really want to see how they measure the “cognitive performance” of an LLM…
I also want to know what part of problem solving exists outside of these tests, as it’s actually a core component of cognitive testing.
Seems like someone made shit up, I’m guessing using an LLM…