How is AI ANY different than access to a super smart human? Why do we hold AI to a higher standard than the smartest humans in the world?

It's a fair question. Any sensible person will take all advice with a grain of salt, as all sources can be fallible.

The first problem is that there are many people out there who aren't sensible : ) Dangerous levels of cognitive outsourcing are already widespread, but dolts will be dolts.

The big problem that is more specific to the LLMs is that if, say, you're considering the advice to a learned person you know, there is usually (at least hopefully) some sort of correlation between their expertise on any given statement and their confidence in that same statement, which they are likely to append to what they say. By contrast the bots are generally knowledgeable, but REALLY suffer from Dunning-Kruger. I wouldn't expect a doctor friend to say "I absolutely know this drug and everything about it, and I've read the studies, and I assure you it's entirely harmless in your situation", if it's something that is in fact toxic. But an LLM? Sure. The slippery-slope danger that is much worse for bots than for humans isn't getting sucked in by the statement, but getting sucked in by the misplaced confidence. Especially if it has been right several times in a row and you let your guard down.

Jim