Right now, humans still have enough practice thinking to point out the errors, but what happens when humanity becomes increasingly dependent on LLMs to do this

Right now, humans still have enough practice thinking to point out the errors, but what happens when humanity becomes increasingly dependent on LLMs to do this thinking?