Eight wheel segments

Inference of Opinion is where Alexa and Taylor pick a take on the week's AI news and argue it.

Token for Granted covers pricing, tokenomics, and the quiet ways AI bills show up in production.

Worth the Weights? asks whether a model, fine-tune, or architecture change is actually worth shipping.

Ship Happens is the segment about launches, incidents, and what broke (or shipped) in the wild this week.

Context Clues digs into RAG, memory, and how systems actually use context — not the slideware version.

Hallucination Station is where confident wrong answers get caught, named, and dissected.

Model Behavior looks at how models act in products: refusals, tools, agents, and surprising habits.

Human in the Loop is about the people still in the system — review, ops, and where judgment still wins.