“You cannot draw a border around the mystery, put on neat handles that let you use the Mysterious Thing without really understanding it—like my attempt to make the possibility that life is meaningless cancel out of an expected utility formula. You can’t pick up the gap and manipulate it.”
Bullshit. You’ve been doing exactly that for the last 10 years.
The ‘Eliezer 1996 algorithm’ safely recursively self-improved to the ‘Eliezer 2008 algorithm’, even though the scientific concepts represented in the original Eliezer algorithm were only vague, general and qualitative.
Furthermore, to communicate any scientific result (ie via speech, text etc)implies that that result has a qualitative high-level (ontological) representation that fully expresses all the required knowledge just fine.
Humans were reasoning before Bayes. Which proves that Bayes is incomplete.
“You cannot draw a border around the mystery, put on neat handles that let you use the Mysterious Thing without really understanding it—like my attempt to make the possibility that life is meaningless cancel out of an expected utility formula. You can’t pick up the gap and manipulate it.”
Bullshit. You’ve been doing exactly that for the last 10 years.
The ‘Eliezer 1996 algorithm’ safely recursively self-improved to the ‘Eliezer 2008 algorithm’, even though the scientific concepts represented in the original Eliezer algorithm were only vague, general and qualitative.
Furthermore, to communicate any scientific result (ie via speech, text etc)implies that that result has a qualitative high-level (ontological) representation that fully expresses all the required knowledge just fine.
Humans were reasoning before Bayes. Which proves that Bayes is incomplete.