Using GPT-3.5 to render visual interfaces within a conversation

I don't understand what is happening here. How is working with the audio for example?

I’m not sure if these are real examples. If they are tho, then it might be using agents that do have that ability.

I think the larger point of the article is to show the user what an agents are doing in a visually understandable way.

Using GPT-3.5 to render visual interfaces within a conversation (symphony.run)