The Dunning–Kruger amplifying loop

A new cognitive pattern is emerging:

1. User forms a shallow belief.
2. User asks LLM to “explain” or “justify” it.
3. LLM outputs something confident and coherent.
4. User interprets coherence as correctness.
5. User's confidence rises.
6. User stops questioning.
7. User repeats the cycle.
8. The extremity of conviction increases.

People with zero literacy in a domain suddenly speak like authorities, because the model gives them the linguistic uniform of authority.

This is an amplified Dunning–Kruger engine. And it will compound.

The world needs tools to guide people through this arc. It needs to support intellectual mirrors with introspective ones.