It’s an illusion because we speak at a higher wpm rate than we type and models work better with more detailed prompts. Unfortunately you can’t copy paste into a spoken conversation or easily provide links to other relevant data