ohyesrobot.ordoliberal.com

agi

Original: agi on Saturday Morning Breakfast Cereal

Transcript

Panel 1:
Student (off-panel/from audience): Professor, what's the difference between artificial intelligence and artificial general intelligence?

Panel 2:
Professor (a bearded man in glasses, holding chalk): Allow me to explain via the medium of bigotry.

Panel 3:
Professor (at a chalkboard): An artificial intelligence is a machine that is capable of racism, but only because it has been fed biased training data. Its racism may be patchy. Entire landscapes of prejudice may escape its notice.
Chalkboard (drawing of a robot with a speech bubble): I only have American chauvinism

Panel 4:
Professor: An artificial GENERAL intelligence has the ability to assess novel situations in a prejudiced manner - to perceive and react racistly due to sincerely held doucheyness in its model of reality.
Chalkboard (robot speech bubble): I even hate imaginary races!

Panel 5:
Professor: Maybe we shouldn't try to duplicate the human mind, just eliminate it.
Professor: You are now ready to learn AI ethics.

Votey:
Professor (with a hand-drawn robot head sketched beneath the speech bubble): I don't use equations. Just robot heads saying stuff.

Alt text

A five-panel SMBC comic. A bearded professor in glasses lectures. Panel 1: a student in the audience asks, "Professor, what's the difference between artificial intelligence and artificial general intelligence?" Panel 2: the professor, holding chalk, says, "Allow me to explain via the medium of bigotry." Panel 3: standing at a chalkboard that shows a robot saying "I only have American chauvinism," he says an artificial intelligence is a machine capable of racism only because it was fed biased training data; its racism may be patchy, and entire landscapes of prejudice may escape its notice. Panel 4: the chalkboard robot now says "I even hate imaginary races!" as he explains an artificial GENERAL intelligence can assess novel situations in a prejudiced way and react racistly due to sincerely held douchiness in its model of reality. Panel 5: he concludes, "Maybe we shouldn't try to duplicate the human mind, just eliminate it. You are now ready to learn AI ethics." Votey: a sketched robot head sits below the professor's speech bubble as he admits, "I don't use equations. Just robot heads saying stuff."

Transcribed by Claude Opus 4.8.