If I create an algorithm that is supposed to select and interpret salacious or evil things, it will do that.
It’s like saying Google’s search results are unbiased, but every algorithmic tweak is done by biased humans to create their end result.
And it’s all “algorithms.”
[[but every algorithmic tweak is done by biased humans to create their end result.]]
In otherwords- the experiments are ‘intelligently designed and influenced’ - this is why ‘evolution models’ fail to be truly neutral and random- because intelligent designers ALWAYS influence the outcome in one way or another- it’s unavoidable
Anything that discriminates is by definition biased.