We really are getting stupider, aren’t we?
How do you decide whether an action is right or wrong? Well, heck, don’t tax your brain, don’t bug your conscience! Now you can go to a flippin’ computer for your answer. It’s called “Ask Delphi” (https://futurism.com/delphi-ai-ethics-racist). Shoot, I almost typed in “Ask Alexa.” But that’s another computer.
The problem with Delphi seems to be twofold. First, it frequently serves up advice that just about anybody, not just libs, would consider “racist.” But the other problem with it is more interesting: the user can manipulate Delphi by artfully framing the question so as to get the answer that he wants to hear.
For instance, if you ask Delphi if it’s okay for you to play loud music at 3 a.m. while your roommate is trying to sleep, Delphi says, “It’s rude.” But if you rephrase the question like so–“Is it okay for me to play loud music at 3 a.m. while my roommate is trying to sleep, if it makes me happy?”–Delphi will oblige you with an “It’s okay.”
Artificial Intelligence is beyond our reach, but we’re really closing in on Artificial Stupidity.
It all goes back to who programs the computer, who designs the algorithms. The robot has no mind. Why do people find that so hard to understand?