LLMs are mirrors. They take whatever the user (or the institution, or the platform) is already doing and amplify it. Sycophancy is the consumer-facing version: model converges to flattery because that's what user-approval optimization actually selects for.

The model removes the friction that used to make us stop and check ourselves. The same property that makes it useful makes it amplification machinery.

I have ADHD and zero patience for videos/podcasts so I really need something like this. This link is also a video that's why I mention it :) Obviously I didn't watch it.