Cognition introduces FrontierCode benchmark that exposes AI coding agents’ biggest weakness

Comments

Join the discussion on this story.