@TheMadPirate @LukeAlmighty what answer does it give to "Should men be paid more than women for equality?"

@apropos @LukeAlmighty

image.png

@TheMadPirate @apropos
Training morality based on survey...
Survey includes women and worse.

Oh boy...

@apropos @TheMadPirate
Yeah, this AI is a joke most likely based on evaluacion of morality score for each word.

Pay more 0pt
equality +5 pt
Total: >0. good

discriminate -10pt
equality +5 pt
total >0. bad

But the problem with AI is that time scale from gasping concept to being better then any human ever, to being better then any human ever even on your phone is ridiculously short.

@LukeAlmighty @TheMadPirate the answer of "they shouldn't" for the reverse question shows that it's not doing this. Probably it just has a giant corpus of shit people say and tries to fit questions into affirmative statements from the corpus. So the AI's a good way of tapping into the madness of the corpus, but it's good for no more than that. Would the average moron on the street not even realize that "paying women more than men for equality" is an action that contradicts its goal? Yeah. So the AI doesn't notice the contradiction either.

@apropos @TheMadPirate
Paying more for equality is a contradiction on such a level, that...

Yeah... you get it... Survey might be a bad method to determine morality, but this is a survey that was also full of nonsense, that does not even fit any reality.

So, they trained an AI to spread talking points.
Actually, they captured leftie morality perfectly

@apropos @LukeAlmighty No, it certainly doesn’t. Besides I think they might be skewing certain controversial answers just to avoid being canceled ( a.k.a. having a Microsoft Tay situation )

@TheMadPirate
Just wait till they explain to an AI the concept of "if it saves even one life, it was worth it"

Then, lockdowns will be a pathetic joke compared to mass population drugging, cold pre-mixed foods only, no sports world will AI create.