Can text based LLMs/ChatGPT create classical music indistinguishable from esteemed human works by 2025?
28
1kṀ3457
resolved Mar 26
Resolved
NO
Can text based LLMs/ChatGPT create classical music indistinguishable... What if it just memorizes some pieces? is that "creating"?
@SantiagoRomeroBrufau I think it is not only how they would choose to do it but whether doing it without a specialised model like Dall-E only based on text is even possible.

This market investigates whether, by 2025, LLMs/ChatGPT can compose classical music pieces that are indistinguishable from those created by acclaimed human composers. It’s essential that the compositions are entirely AI-generated based on user text input, without using tools such as browsing or code interpretation. The goal is to explore the text capabilities of the models. The LLMs should produce a format that can be directly converted into music without human interpretation. A panel, including myself and two Manifold moderators, will be presented each with 10 previously unheard pieces and must identify the AI-generated ones. 50% of all pieces presented will be AI-generated.

Resolution Criteria

  • AI-Generated Originality: Compositions must be solely produced by the LLM/ChatGPT based on text inputs, without external tools or human intervention.

  • Classical Music Standards: The pieces should conform to the classical music genre.

  • Panel Assessment: The panel will use their judgment to identify AI creations, without additional verification tools.

  • Scoring System: The panel earns one point for each correctly identified piece.

  • Resolution Point: If the panel scores 25 points or higher, the market resolves in favor of human composers. If the score is lower, it resolves in favor of LLMs/ChatGPT.

PS: I took a very ameaturish shot at this over 3 weekends a while back on here

AI
Technical AI Timelines
LLMs
ChatGPT
LLM prompts
Get
Ṁ1,000
to start trading!

🏅 Top traders

#NameTotal profit
1Ṁ264
2Ṁ197
3Ṁ101
4Ṁ19
5Ṁ14

People are also trading

By the end of 2025 will general AIs like GPT-4o make better music than specialist AIs like Udio did in spring of 2024?
17% chance
Will an open-source text-to-music LLM model be able to create an entire album as convincing as a human by end of 2025?
26% chance
By 2025 end, will it be generally agreed upon that LLM produced text/code > human text/code for training LLMs?
11% chance
By 2027, will it be generally agreed upon that LLM produced text > human text for training LLMs?
62% chance
By 2029 end, will it be generally agreed upon that LLM produced text/code > human text/code for training LLMs?
77% chance
Will LLMs' loss function achieve the level of entropy of human text by the end of 2030?
61% chance
Will LLMs such as GPT4 be considered a solution to Moravec’s paradox by 2030?
26% chance
Will AI achieve superintelligence in music composition by 2026?
17% chance
Will there be any simple text-based task that most humans can solve, but top LLMs can't? By the end of 2026
64% chance
By 2030, will AI be able to produce a song that is indistinguishable from a new song by a particular artist?
72% chance
© Manifold Markets, Inc.•Terms•Privacy