This is a key starting point. We have to spend a good deal of time understanding how our minds adopt tools (Language, numbers) in order to work better.
There is a big difference between complementary cognitive artifacts, and competitive ones. For example, language complements and enhances the mind, but I’m not so sure that Neuralink will. My initial impression is that Brain-Computer interfaces will work like many of our smart phone applications, only more intimately, further worsening our natural capacities for memory, spatial navigation, and relevance realization. An exploration of complementary psychotechnologies can be found here:
This work will provide some grounding for understanding what a human ought to focus on, once AI becomes a master of even unbounded problem solving.