Due to its random uniform distribution of the severity of its errors (meaning: it’s just as likely to make miniscule errors as it is to make massive deadly errors that even the dumbest of human would not do), AI is best used in situations were even the worst of errors are not a problem.
The American Military and American Administration are clearly groups for whom even the worst of errors in targetting when bombing other countries are not a problem (schools full of children get tripple-tapped once in a while is fine for them), hence AI is from their point of view fine to use for that without supervision.
Comments
Due to its random uniform distribution of the severity of its errors (meaning: it’s just as likely to make miniscule errors as it is to make massive deadly errors that even the dumbest of human would not do), AI is best used in situations were even the worst of errors are not a problem.
The American Military and American Administration are clearly groups for whom even the worst of errors in targetting when bombing other countries are not a problem (schools full of children get tripple-tapped once in a while is fine for them), hence AI is from their point of view fine to use for that without supervision.
Well it’s a good thing they call it SI now I guess.
Sign in to comment. Sign in