LLMs discern patterns in words used together, by both experts and the general public in training data. Then they can look at patterns in context, meaning, emotional tone, etc… when you ask it a question it looks at its training data, looks at you and then can respond remarkably appropriately

Did you ask it what is the weather in YOUR SPECIFIC LOCATION?

I asked mine and it got it right.

Oh because everything else on the internet that Google exposes to you is true and no discernment is required there at all… ChatGPT has been trained on terabytes of data and can see true patterns most humans can’t

To help me overthrow the patriarchy