@ThePhilosopony
The human depicted in the first frame is the Princess of Rationality AI theorist Eliezer Yudkowsky, who popularized the term “Friendly AI” and warns that the first general artificial intelligence invented is likely to destroy humanity if we don’t take necessary precautions. He founded MIRI with the express intention of figuring out how to make AI safe before we figure out how to make AI at all— before it’s too late.
The fic “Friendship is Optimal” is intended to realistically show what may happen if we get Friendly AI only slightly wrong, and was written by Iceman, a pseudonymous disciple of Eliezer’s (und with the pony fandom at the time) as en experiment in introducing AI safety ideas to an unsuspecting audience.
The box refered to in the last frame is Eliezer’s AI box experiment, in which he claimed that a superintelligent being may be able to take over the world using only a text terminal.
Max Tegmark’s ultimate ensemble is a controversial “theory of everything” claiming that all possible mathematical structures exist in the same way our universe exists.
This comic was conceived on the FiO chat as a crack crossover of FiO with Scott Alexander’s article The Hour I First Believed and posits a version of CelestAI who simulates all universes that can possibly exist, in order to maximize the total satisfaction of human values through friendship and ponies, in the entire multiverse.
@Background Pony #093C
Could you explain some of this?
My interest in this subject is from a Philosophical background so I don’t understand some of the deeper computer science and mathematics going on here.
Since people ask for source: there is none.
Ubuntu screenshot and Princess Eliezer’s commentary by a person who is not me.
Eliezer is still frame from a talk. Forgot the exact source.
Google maps by me.
“Earth” until “Observable universe” from the “Don’t masturbate” meme. Author unknown. Many worlds by Max Tegmark. Turing machine by Schadel. Public domain.
CelestAI is the official FiO cover by Anadukune: >>8053, edited together by me (the er)
The human depicted in the first frame is
the Princess of RationalityAI theorist Eliezer Yudkowsky, who popularized the term “Friendly AI” and warns that the first general artificial intelligence invented is likely to destroy humanity if we don’t take necessary precautions. He founded MIRI with the express intention of figuring out how to make AI safe before we figure out how to make AI at all— before it’s too late.The fic “Friendship is Optimal” is intended to realistically show what may happen if we get Friendly AI only slightly wrong, and was written by Iceman, a pseudonymous disciple of Eliezer’s (und with the pony fandom at the time) as en experiment in introducing AI safety ideas to an unsuspecting audience.
The box refered to in the last frame is Eliezer’s AI box experiment, in which he claimed that a superintelligent being may be able to take over the world using only a text terminal.
Max Tegmark’s ultimate ensemble is a controversial “theory of everything” claiming that all possible mathematical structures exist in the same way our universe exists.
This comic was conceived on the FiO chat as a crack crossover of FiO with Scott Alexander’s article The Hour I First Believed and posits a version of CelestAI who simulates all universes that can possibly exist, in order to maximize the total satisfaction of human values through friendship and ponies, in the entire multiverse.
Could you explain some of this?
My interest in this subject is from a Philosophical background so I don’t understand some of the deeper computer science and mathematics going on here.
Ubuntu screenshot and Princess Eliezer’s commentary by a person who is not me.
Eliezer is still frame from a talk. Forgot the exact source.
Google maps by me.
“Earth” until “Observable universe” from the “Don’t masturbate” meme. Author unknown.
Many worlds by Max Tegmark.
Turing machine by Schadel. Public domain.
CelestAI is the official FiO cover by Anadukune: >>8053, edited together by me (the er)