I had a human one of them. I spent more time interacting with it and dealing with misunderstandings and corrections than I gained from not solving the problems myself.

As pointed out elsewhere I was gaslit into changing the subject by an LLM the other day and then it got stuff wrong. End game is it created more work than not using it.

And that’s what we need to be careful of. The promise versus benefits of technology are not always obviously swayed towards the benefit end.