Is that what the article is saying though?

Currently revising for a psychology exam on human-automation interaction and from what they've taught us, this makes a lot of sense because automating skills (like idea generation and decision-making) makes your own skills degrade from lack of use. There's decades of evidence on pilots' use of autopilot saying that the ones who lack confidence in their skills use it most. So it's a vicious cycle. You lack confidence in your ability to make a decision or generate good ideas, so you unload it onto a bot (an unexplainable one whose methods, rationale and evidence is obscured from you) and your skills in the whole decision-making process of researching, evaluating and synthesising evidence begin to decay. And this increases your reliance on the AI to do these things for you. .