I've downvoted, because it's not clear to me why or how this would be useful. Please consider adding a sentence or two about motivation. As part of that, I think it needs to be made clear how this is a music question rather than simply a computer technology question that happens to revolve around sound.

Related post: Is there any objective genre-like categorization of musical works?

Are you sure this is a question about music practice and/or theory? Seems to be about computer science to me.

This entire question, & comments underneath, fall into the philosophical trap of argument from ignorance. In short, "Because this hasn't yet been proven to be impossible, it should just work, right? If such a tool existed new genres could be developed and explored much faster, perhaps in days." Errmmm… no.

I vtc, as it's not even possible to pigeonhole certain songs, so how would anyone - or any computer program do any better?

@user87976 isn't that how every genre with "fusion" in its name works? "Progressive" involves that a lot, too.

@user87976 I don't really see why it would be more difficult than generating music in general. The problem is that generating music itself is more or less an unsolved problem and the existing research seems to be more towards generating copyright-free background music than creating something new. The other problem is that genres can't really be defined as simply as OP thinks so nobody who understands it is even trying.

Hi Steve - please re-read our tour and How to Ask pages - You need to have a clearly focused, single question, and search the site for others before posting. 1 and 2 are explicitly off topic here, and 3 is provably unanswerable - we already have questions on how to define genre (hint - it really can't be done - everyone disagrees on every single genre)

I suspected that this question was too large to be fully answerable here due to our current lack of understanding of what music (and more specifically what genres and fused genres) are. Based on most of the comments here the 59 people who have viewed this question believe we are still, metaphorically speaking, clanging rocks together in a cave. So I have my answer; you are aware of no such AI. Thank you all for attempting to answer this question, I know it was a challenging one . . .

However, due to the brevity of this discussion and my own study, I'm convinced that much more research into this than would be indicated here has been done. There is a lot going on at MIT, Berkley and elsewhere. And just because we are unaware of a certain technology today does not mean that it doesn't exist or isn't being seriously worked on. This problem will solved, it's just a matter of time and effort. That said, I do not know how to re-phrase this question to be acceptable to this site. If someone knows how, I encourage you to do it, otherwise it will remain closed. Thanks again.

Your extrapolation is as incorrect as your initial assumption. You are asking "How can I build a bridge from Sydney to Auckland?" & after being told it is impossible with current technology your conclusion is that humans have not yet learned to build bridges.

@Steve the question is totally answerable and has already been answered. It's not exactly what you asked for, because the question is based on total misunderstanding how genres work.

@ojs Perhaps we all misunderstand (or don't understand) how genres work?

@Steve Citing en.wikipedia.org/wiki/Music_genre, "musical genre is a set of musical events (real or possible) whose course is governed by a definite set of socially accepted rules", where a musical event be defined as "any type of activity performed around any type of event involving sound" sounds pretty accurate to me. Note that it's about socially accepted rules, not just the sounds.

@Tetsujin. No. I am concluding that YOU are concluding that it is currently not possible to develop an AI that can fuse 2 different genres and thereby produce a new genre . . . and you may be right :) Remember, I'm simply asking the question.

Your logical fallacy is now also advancing towards an argument from incredulity. This, again, is because you do not understand the exact nature & complexity of the question. "Surely someone must be working on this" is a logical fallacy. As no-one can adequately define any specific genre makes this a logical dead-end.

@ojs Where did this idea that I am talking about sounds come from? I'm talking about musical genres; NOT sounds.

@Steve The question as it is written says "take audio samples". If it is not about sound, I don't know what it is.

@Tetsujin Ahhhh, now we may be getting somewhere. There is no logical fallacy on my part here. If in the course of developing such an AI that a mathematically, or waveform or emotional crisp definition of a genre needs to be produced . . . then so be it! I see no problem with solving 2 problems in parallel or in serial. It's done all the time.

That's just using more words to re-state a starting point that is unachievable. I'm done now. I'm afraid lack of knowledge as to what forms the basis of your question is not going to be helped by further recursion.

@Steve technically the audio samples are representations of sounds, which may or may not be music depending on the sound they represent and your definition of "music". Reproducing the sound and your perception of it what makes them music, and genres are much higher level of (subjective) interpretation.

@ojs, Respectfully I still don't think you understand. There is no "may or may not be music" involved here in the least. I am talking about music; full stop. And it's not just my personal perception or interpretation only but that of society's. The audio samples within a genre I am talking about would be universally accepted as true music; taken from a Pandora playlist, a greatest hits album, a highly acclaimed movie soundtrack or similar. Why on earth would I be interested in noise or non-music?? And more importantly, what did I say to lead you to this misunderstanding??

@Steve it's not misunderstanding, it's that you have the concepts of genre, audio sample, music, etc all wrong. Just go and read about the basics of signal processing and musicology and you'll understand.

@ojs, this is going nowhere. I have read up on these subjects and they have nothing to do with you wrongly thinking that I'm talking about non-music when in fact I'm talking about true music. as explained above. The bottom line here is that you are unaware of such and AI. And guess what? You're in good company, neither am I. That is why I posted the question in the first place. I'm just amazed at how easily you have misunderstood what I have been saying from the beginning. Perhaps if you re-read from the top you will see where/how you missed the message.

The essential problem here is that you don't understand the complexity of the question. This is not in itself a bad thing, any beginner in a subject will hit the same wall. The problem comes that you don't understand that you don't understand & are refusing to accept people telling you you don't understand. Please read up on the Four Stages of Competence & the Dunning Kruger effect. You have to accept what people are telling you here.

@Tetsujin. What can I say to convince you that I do understand the complexity here. But that doesn't preclude us from discussing it as adults and recognizing that today's complexities are tomorrow's reality. And yes you can quote me on that. You don't seem able to understand that we, as a race, are on a technological spectrum. At one point of this spectrum digital music was science fiction. Now it's commonplace. Was it you that said ""Surely someone must be working on this" is a logical fallacy. " ? How can you say that is a fallacy? Are you all-aware of what MIT is doing? . . .

And no, I do not have to accept all what people are telling me here. In order to accept something, it must pass a reasonable test of plausibility. ojs' wrong understanding of me talking about non-music is far from plausible. I can never accept that falsehood. Would somebody please message him and tell him I am talking about true music. Reference Pandora, greatest hits albums and award winning soundtracks.

I don't see anywhere he misconstrued you, but several times you misconstrued what he was trying to explain to you. This is the problem with your not understanding what it is you don't understand.

Comparing "someone, somewhere must be working on this" to the advent of digital music is another false logic, whether you like it or not. I'm really done now. Choose to accept this or not.