LeCun seems to believe that only "human-level AI" could create huge, possibly catastrophic unintended consequences. That's not the case. Perfectly stupid social

LeCun seems to believe that only "human-level AI" could create huge, possibly catastrophic unintended consequences. That's not the case. Perfectly stupid social network algorithms have already huge detrimental effects on the mental health of millions of people. Perfectly stupid social network algorithms have big, serious and unexpected political consequences around the world, and can provably make or break elections for instance. Implying that we need "human-level AI" to create a catastrophe is not merely short-sighted, in the light of what we already know it's either really naive, or a deliberate act of misinformation.