Prefer a chat interface with context about you and your work?
How optimal is word recognition under multimodal uncertainty?