Yuuko faces the music for her lax homework habits.
Hakase goes for a walk and gets scared by dogs. Yuuko and Mio try to help and get bitten.
(Several other sketches feature rare characters that I am too lazy to introduce at this point.)
It kidnaps the president of the World Chess Federation to get him to revise the official rules of chess to say <chessmaster always wins>.
_Superintelligence: Paths, Dangers, Strategies_
This is the best single source for the memes about AI turning the universe into paperclips.
Sadly, if you've already read a bunch of LessWrong, this does not have better arguments, just a different flavor of narrator.
Mio finalizes her manga for submission. The others help.
Nano overheats when she sees the subject matter.
Yuuko spills ink on the pages.
Mai draws whatever she wants in the background.
Some help.
Hakase plays hide and seek, but hides too well.
Yuuko helps Mio train for the high jump.
The girls go to the zoo to see white tigers.
Hakase fails to bring in the laundry.
The girls have a run of bad luck.
Basically, the LessWrong arguments are not always careful to distinguish between Seed AIs, Superintelligent AIs, Sovereign AIs, Perfect Bayesian Reasoners, Optimizers, Optimized AIs, and probably several other kinds.
I expect the arguments are full of type errors.
Hakase and Nano count sheep to help fall asleep. It's super effective.
Yuuko and Mai go trick-or-treating to get free candy. The only problem with this plan is that Halloween was last week.
Hakase makes a biscuit-powered robot, mostly to get attention.
Hakase befriends a crow.
The girls try to build a house of cards.
Nakamuram tries to catch Nano, but only catches other people instead.
Yuuko tries not to respond to Mai's joke setups.