OpenAI Whisper Quiz

Answers and explanations appear as you take the quiz.

1. What is the main purpose of OpenAI's Whisper?
   • Automatic speech recognition: transcribing and translating audio
   • Generating images from text
   • Writing computer code
   • Creating 3D animations
2. What was true of the language balance in Whisper’s training audio?
   • Every language contributed the same number of hours
   • The dataset was English-heavy, with much less audio for most languages
   • The dataset contained only English audio
   • Most data was studio-recorded speech in one language
3. What does Whisper use to condition the decoder on the requested task and language?
   • A separate camera feed
   • A hand-written transcript of the answer
   • Special tokens in the decoder prompt
   • A unique model checkpoint for every utterance
4. Which capability does Whisper NOT natively provide?
   • Translating speech from other languages into English
   • Detecting which language is being spoken
   • Adding timestamps to a transcript
   • Generating realistic video from a prompt
5. Which long-form transcription problems does OpenAI’s paper describe?
   • Repetition loops and missing words near segment edges
   • Perfect diarization of every speaker
   • Automatic correction of every proper noun
   • Guaranteed removal of background music
6. What is a known weakness of Whisper to watch for?
   • It cannot process any English audio
   • It can occasionally 'hallucinate' text that was not spoken
   • It only outputs audio, never text
   • It refuses to add punctuation
7. Approximately what share of Whisper’s training data was English audio paired with English transcripts?
   • About 18%
   • About 1%
   • About 65%
   • About 100%
8. What use does OpenAI caution against when transcription errors could have serious consequences?
   • Noncommercial experimentation with public samples
   • Adding timestamps to a transcript
   • Using the model to identify a language
   • High-risk decision-making without appropriate verification