Question: Explain why turning a spoken recording into written text is an example of rich-output learning. Describe the input, the output, and why this task fits an end-to-end learning approach.

Sample answer: Speech-to-text is a rich-output learning problem because the model takes in a long audio signal and produces a detailed text transcript. The output is not a single label or number; it can contain many words and complete sentences. That makes it much richer than tasks that predict only one class.

This is a good example of end-to-end learning because the system can be trained directly from paired examples of audio and the correct transcript. With enough labeled pairs, the model can learn to map the sound pattern to the written words without needing every step to be manually designed.

Key points:

• Identifies the input as audio.
• Identifies the output as a transcript.
• Explains that the output contains many words, not one simple value.
• Connects the task to end-to-end training with labeled input-output pairs.

Rubric: A strong answer should name the audio input and text output, explain why the output is rich rather than simple, and describe how labeled audio-transcript pairs support end-to-end learning.