AI and Music Lecture

Jul 15, 2024

AI and Music Lecture

Introduction to AI in Music

  • Speaker discusses advancements in AI-generated music.
  • Introduction to Jacob Ker’s song vs. an AI-generated song.
  • Speaker notes how realistic AI music sounds.

Identification of AI Music

  • Speaker shares personal anecdote about playing AI music at home.
    • Dylan, a 10th grader, could instantly recognize the AI-generated songs.
    • Test conducted with 10 AI songs vs. non-AI songs with similar results.
  • Dylan senses something “weird” in AI-generated voices.
  • Other kids (Dylan's peers) struggled to identify AI vs non-AI music.

AI Compositions

  • Examples of AI-generated songs shared via email to speaker.
    • Example 1: Sad beach song.
    • Example 2: Metal band from Belarus using fully AI for their music.
  • Utilization of the "udio" program for AI music generation.
    • Example 3: Movie score in the style of Ennio Morricone.

Future of AI in Music

  • AI is expected to be widely used in the future for music composition.
  • Speaker suggests people will bypass traditional composers in favor of AI.
  • Majority of listeners unable to discern AI from real compositions.

Personal Experiences

  • Speaker shares another incident with daughter, Leila, recognizing AI music.
  • Family kids seem more adept at identifying AI music due to a certain 'echo' or 'sound'.

AI Music Prompts

  • Prompts used to generate AI music: examples include various genres and style-specific cues.
    • Example: prompt for chill music including genres like Chillstep, House, HipHop.

General Observations

  • Digital processing of human voice makes AI indistinguishable from real in some cases.
  • AI might replace human musicians, benefiting companies who develop AI tools.
  • Speaker notes using AI for interactive improvisation could be an interesting use-case.
    • Use of AI in live performance settings suggested.

Ethical and Economic Concerns

  • Discussion about ownership and profit: who gets paid for AI-generated works?
  • AI on training data from existing human creations—ethical implications mentioned.
  • Congressional testimony by speaker reveals slow legislative response.

Conclusion

  • AI is at an inflection point in music industry; curiosity about audience's ability to recognize AI music.
  • Speaker asks for feedback from the audience and encourages engagement (subscribes/comments).