You mean that we should do linear regression between LLM activities and brains? 🤡

More seriously, I find those views refreshing and important, but they might have similar issues. As Cisek brilliantly put:

We are like magicians wondering: 'Maybe there really is magic?'