[Thoughts] Carol's PyBeach Keynote and AI's Polarization Problem

Finishing Carol Willing's keynote from PyBeach.

An emotional moment for me was of her showing the juxtaposition of how AI can create bias by asking for a image of a person with late stage Alzheimers vs a picture of her own mother in the same stage and how polarizingly different these things are.

It reminded me that the internet and AI in many ways, want to generate the most "whatever emotion" response as possible.

Reminders:

• ChatGPT (or whatever you use) is not a human (except when it copy/pastes content from humans)
• These companies have an opportunity to give you content that is free of bias, and the fact they don't is an intentional decision made by humans.
• You are smart but marketing will leverage how smart you are against you.
• Any time you "Feel a way" about a result, you need to examine that feeling and determine was it engineered by the LLM.

This isn't me slandering it, it's me having used these exact strategies against other humans in a past life when I was in marketing. I hope I brought on feelings of mistrust towards these companies and their intention of making trillions off you and your feelings.

I'm just as bad as them.

-- Anyway! I highly encourage you to checkout Carol's talk! I was fantastic with a great blend of inspiration and technical content.

https://www.youtube.com/watch?v=Vz_-NQG8Jcg&list=PLwdA5WU2fAAvCIOjgnYDA7NrGJ8iFh97t&index=1