A quiet day on the lab front — nothing new from the model companies in the last 24 hours that wasn’t already covered. Today’s list is two pieces from individual authors: Armin Ronacher on how LLMs quietly steer independent builders toward the same ideas, and Simon Willison’s pelican-grid comparison of GPT-6 Astra against the GPT-5.6 family across reasoning levels and cost.
Worth your time
- Latent Powers — Armin Ronacher
- The Pelican comparison grid for Astra is pretty interesting — Simon Willison