Michelle Pellon's Blog

On Arendt, Scheler, and the bodhisattva vow — keeping the higher things alive in an age when machines can simulate nearly everything that matters.

AI-generated code is eroding the shared mental models teams need to maintain their systems.

A response to Cory Doctorow arguing that conflating financial bubble dynamics with AI capability assessment leads to flawed conclusions.

How a 28.6% accuracy week revealed my NFL prediction model was collecting weather and rest data but completely ignoring it.

Build a sophisticated NFL prediction model achieving 57% accuracy using ELO ratings, DuckDB, and dbt.

Claude Skills, packaged as a governed, repo-first library, are the practical path to scalable, cross-domain human-AI collaboration in the Enterprise.

Exploring the risks of chatbot interactions amplifying delusional beliefs and the urgent need for a Belief-Amplification Index to measure and mitigate AI-facilitated delusional ideation in vulnerable populations.

AI agents promise convenience, but at what cost?