Review pans GPT-6.2 Astra for poor usability and circular reasoning on long tasks
A review of GPT-6.2 Astra criticizes the model for lacking awareness of knowledge gaps, producing hard-to-parse outputs, and running in circles on long tasks.
Read the full story at Techmeme ↗