"Scientists Built an AI to Give Ethical Advice, But It Turned Out Super Racist"
Pretty much. We tricked millions into loving statistics and higher level math. All it took was painting a picture about how such mechanisms seem so human. As if being humanoid is somehow intrinsically important.
But, any computation fits the definition of some level of intelligence. People will decry that it is or is not AI until every productive human activity and factulty is replaced and optimized by it.
+bonifartius 𒂼𒄄
@lupyuen "nothing new in the west"
why do people expect a moral AI if the training data is from our amoral societies. it's just a mirror.
it is really time for the next AI winter (which should be really long please) already. the stuff we have isn't more "intelligent" than eliza.
So many chatbots end up this way... They could use twitter feedback to train as an adversarial dataset.
@lupyuen turns out we haven’t gone as far past Dr Sbaitso as we thought, just calling it “AI” is a good way to fool investors.