I don’t think “AI” will outsmart us. I think one aim or purpose of Large Language Models is to dumb us down—to condense the complexity of language and thus obfuscate & confuse meaning as a way to degrade communication and social or group intelligence.