I love this: hazmat warning labels for data science. https://very-good-science.github.io/data-hazards/index.html
@yaaps I would accept a warning label like “this data set has been handled by somebody who takes Roko’s Basilisk seriously” tho
"Whatever is right can be achieved through the irresistible power of awakened and informed public opinion. Our object, therefore, is not to inquire whether a thing can be done, but whether it ought to be done, and if it ought to be done, to so exert the forces of publicity that public opinion will compel it to be done."
- William Randolph Hearst
"Furnish the pictures. I'll furnish the war."
I'd argue that Roko's Basilisk is a recurring hazard. Not literally, of course, but the "subject to misuse" and "confirms existing biases" labels don't quite cover what Karl Popper labeled 'historicism'. Asimov called it psychohistory, Marx called it dialectical materialism, Hitler delegated it to Goebbels, America privatized it, the books of Moses created it 3 times despite precautions against it. It's a common habit in politics and religion to tell one's own story in a way that projects the inevitability of success for one's own cause, the futility of resistance, and the outcomes to be anticipated, and this deserves a Basilisk Label
Server run by the main developers of the project It is not focused on any particular niche interest - everyone is welcome as long as you follow our code of conduct!