Séb Krier#519
By 2030, I predict that all AI outputs could be up to 70% in italics, due to the growth rate of sheer self-referential meta-profundity contained in every inference.
Italicisation is currently one of the only known safe methods of mitigating risks such as generalised pomposity and recursive self-referencing.
@nickswan73 @sebkrier I’d take that bet
@nickswan73 @sebkrier I understand that a better mitigation method is to use the html <blink> tag
@nickswan73 Fun fact: italic is based on the handwriting of one guy, Niccolò Niccoli, who was a core part of renaissance book hunting to rediscover and copy lost Roman works
Séb Krier#519By 2030, I predict that all AI outputs could be up to 70% in italics, due to the growth rate of sheer self-referential meta-profundity contained in every inference.
Italicisation is currently one of the only known safe methods of mitigating risks such as generalised pomposity and recursive self-referencing.
