1. Didja accidentally blow through the whole, "We're using our real names" thing on registration? No problem, just send me (Mike) a Conversation message and I'll get you sorted, by which I mean hammered-into-obedient-line because I'm SO about having a lot of individuality-destroying, oppressive shit all over my forum.
    Dismiss Notice
  2. You're only as good as the harshest criticism you're willing to hear.
    Dismiss Notice

AIVA - Houston we have a problem?

Discussion in 'The RedBanned Bar & Grill' started by Alexander Schiborr, Mar 2, 2019.

  1. #1 Alexander Schiborr, Mar 2, 2019
    Last edited: Mar 2, 2019
    So, yeah there was this thread on ViC a while ago where someone posted a link to the page where a company is releasing program which writes music entirerly for you. And I don´t know anything about the details there but I took a listen to a couple of example ranging from very uninteresting to quite better. I found also this example which they say their program did wrote entirely:

    Now any thoughts what do you think about that? In fact I enjoyed the track quite a bit and while having constanly in my head that there is something at times strange, it might be suggestive feeling because I know that it comes from an AI. if somebody wouldn´t tell me that a computer program wrote that (which I am not sure if that is entirely here the case really) I wouldn´t think that at all.

    While I don´t think it will replace excellent composers I feel that it might replace a dozens of people out there in the near future when this is really here the case that this AI can produce music like that.

    Any thoughts? Is that a thing where we should be worried about or just laughing? I am not sure, I am between finding that a bit irritating at the moment to put that mild.

    EDIT: I just played that piece to my wife and she said something interesting: While she was surprised that this is created by AI, she said that emotionally there was nothing what that track strikes for her. You know she is no composer or musician, she is the typical (nearly) consument and I found that statement quite important.
  2. Try reducing this to two-handed piano to find out what it really is. Other than boring, I mean.
    Rohann van Rensburg likes this.
  3. Humans love to replace themselves with machines.

    As far as I'm concerned if a computer wrote the music then it's not art, the computer program might be considered art though :p
    Rohann van Rensburg likes this.
  4. I remember the Aiva Thread on vi-c, I was quite bummed out by the thing because I have to admit that it is a better composer than I am and that created a very weird "why even bother?"-feeling in me. If I was a fulltime composer without high-profile clients I would be scared for sure. This thing is gonna replace some people, and with time the number will increase with increased capabilities of the tools.
    I don't think Hans Zimmer has anything to worry about, because his name alone still carries enough weight to be worth hiring him over an AI for marketing reasosn, no matter how good the AI is.

    I've had a discussion about AIs replacing us yesterday, and a friend said that they're not capable of "true creativity" and creating something fundamentally new that expands the charted possiblity-space instead of just moving within it. Personally I'm neither sure that this is actually true longterm, nor am I sure that more than a fraction of the human composers are capable of that either. I think AI will be able to replace everything and everyone, if we give it enough time.

    I think replacing illustrators and painters will take a fair bit longer than composers, but I'm still scared of the day when it will be able to replace me. I think one of the factors that can keep you in business past the point where you are no longer able to compete on price, time and quality with an AI, is being "more pleasant to work with". Replacing that kind of human element is still a far way off. But it also really depends on the kind of client we are talking about. Ad agencies that need 30 storyboard frames overnight will abandon the human illustrators at the drop of a hat once AI is able to deliver what they want, because it will produce more reliable and predictable outcomes faster, and in that kind of environment it is a plus, where in others it might be a minus.

    If it's neccessary I think I would be capable of switching jobs over to the AI research and development side, but I don't know when to make that call. The longer you wait, the higher the chance that you'll be part of a bigger wave of people whose jobs just got replaced by AI flooding the market.

    A friend of mine just got his PHD and is working on an AI related project, and he said there's currently a 14-day crash course that takes anyone with a PHD (doesn't even have to be computer science related), teaches them a couple of things to get started in working with deep-learning AI systems, and at the end they get about half a dozen job offers from different companies, and the one they are signing with is paying for the course...
    Considering that I already have plenty of non-AI related programming experience it would probably be easy for me to get a job in that field if I wanted to. Bud I'd feel a bit like I'm helping killing off millions of jobs that I don't want to see automated.

    I'll end with a quote from a youtube video called "Humans need not apply":
    "Horses didn't become unemployed - they became unemployable."
    JP Beveraggi likes this.
  5. You know...... it was much better than I thought. It's fractured....most likely they ran it so it was in 3 self contained movements.

    It's also interesting to hear the pieces that must have been one of sources. Sometimes really interesting directions.

    I mean.....it's not without flaws, or "boring" moments. I not arguing that. But it would be in the top third tier of quality if posted at Vi-C.

    One thing for sure..... this program takes more use of counterpoint than most pieces I hear there. For me...... that's a plus.


    But why the "fear" ? I fear lots of technologies, but not this. Why would I ?

    I love writing music, and I love practicing playing instruments. If I have an "ideal" day..... I will work on my own music all day/night.

    The perspective at Vi-C seems only to think in terms of "gigs", and money. It's odd to me. It's almost very odd to me.
    Not that thinking about money is odd. More that I would wager a lot of the most worried already work full time in something else.

    I know the sinking "it's over for me" ....believe me.... I do. But those are stories for another time. I can empathize, but I am suggesting
    if gig's is your only motivation.... your going to burn out at some point anyway if you encounter a "dry spell".

    You know.... instead of just "lecturing" maybe I can say my whole opinion by sharing a conversation I had with a friend of mine.

    We were having a cigar in the park, after drinks and chatting away. He is a successful entrepreneur, and just had his company put on the
    NASDAQ about 6 months prior. (That's a big fucking deal for anyone who does not know.) Anyhow, before he did the start-up thing he was in a
    band that had a pretty good following. They were on a development deal with a major label, and then the whole Napster thing happened, and they got cut...
    so he went back to University, got married, had kids, launched company, now totally rich.

    He no longer plays music, other than 30 minutes on the guitar on Sunday if he can. He sure seemed to miss it.

    I asked him if there was a way for him to put more time into it. (His company was being sold in about 3 months. He was never going to need to work again for money as long as he wasn't a total ass clown with money.)

    I suggested it can be good to set up a day/time with a friend to jam, or a just recording improvisations, or a project. Basically anything that would help from "Life getting the way" .

    Then we kinda got real. He said "You know how many bands are out there. People grossly underestimate how many." Etc...... and
    then then "People should really think about that, because if you are not going to hit your goal, whats the point? It's just wasted time"

    Now I knew the answer to this question so I asked " Do you like to exercise " He says " Yeah, I love it"
    I say " It feels great to be healthy, and take care of yourself. It really clears my mind, and that jog down by the Hudson river is awesome"

    He says " Totally, I love jogging down there too. Every week 4-5 times."

    At the time the Olympics were on. So I asked " Did you see Usain Bolt break that record ?" "That guy is insane" implying he had.

    So I asked " Do think you will ever beat Usain Bolt in a race?" "So why do you jog along the Hudson river when he is clearly always going to be better than you ?"
  6. So if the "AI revolution" doesn't scare you, does that mean you believe we are headed to the Star Trek kind of utopia where everyone is free to devote themselves to the betterment of their crafts without any monetary needs? I would want that to be true but I don't think it is, because systemically nothing is in place to guide us there.

    I'm not sure the jogging/excercise example is 100% the same thing because most people don't excercise to compete, they do it to fend off the ongoing deterioration of their bodies to preserve a greater quality of life for their later parts of life (or to look better). And then there's the "runner's high" thing etc.. I don't think quitting to play guitar has anywhere near the bad longterm consequences for a normal person, that a sedentary and passive lifestyle has.
  7. That's misunderstanding my point. The practice...... the having goals, experience of moving my fingers and making enjoyable (on a good day) sounds... that's my motivation. I guess I have always been attracted to the more "Monk" like musicians. Practice is kind of a meditation or spiritual/intellectual active that I am 100% addicted too. As such, as long as I am in full health the "passive" lifestyle is not on my radar. I see music practice (or creativity full stop) as a form of health, and taking care of myself. A machine posses no threat to the continuation of my practice.

    But see......even here the logic is pretty shaky. I do understand what you are saying. I really do.

    Ok.... let me offer a counterpoint that is outside of my "feelings" .
    There are already a 1000 and 1 composers out there who will score a film for free. It's been that way for years.

    Films can already license thousands of songs for tens of dollars, and make a "mash-up" on the fly. I can make a montage of family photos with soundtrack from my phone within 15 seconds.

    There are still many places that offer "pay-to-play" to aspiring bands. I recall on Vi-C someone posted asking if they could pay to be an intern.

    Ah.....then interns. The whole creative arts world is propped up by interns. So not only will people score a film for free, they will do errands for someone working on a film, just so they can be around and observe.

    People already pass off a patch on Omnisphere, another from say Eric Whitacre Choirs, and a drum loop as a composition....so.......

    The first three to four years out of conservatory most of the gigs I got, if I added up the time spent vs $$ Mcdonalds would have paid me more.

    Have you ever worked on a "Vanity" project? I have. A few.
    Basically someone who will toss 6 figures to make that album they always wanted. (Paris Hilton for example)

    Long story short....... it's never been a wise business path. It's also unlike other fields of business in this regards. It's rare for say someone to say "I'll be your accountant for free so I can gain experience".
    Or pay to be your accountant.

    Music is a natural resource. We can either harness it, or give up.

    (Truthfully and respectfully I would say for a lot of people, myself included, the fear is not just money......but that we will never achieve the admiration we long for.)

    So I guess my take is right from Winston Churchill if AI is the enemy.

    Never give in--never, never, never, never, in nothing great or small, large or petty, never give in except to convictions of honor and good sense. Never yield to force; never yield to the apparently overwhelming might of the enemy.” ― Winston Churchill.
  8. The guy with the NASDAQ company failed because he wanted to do music; he didn't need to do music. If there is anything, any life, any option in any way whatsoever that one could see themselves doing instead of succeeding, then they will never beat people for whom this is not an option. We don't have a choice; we are not psychologically or metaphysically complete without being successful musically. We are not balanced; we are, perhaps, not healthy. But we cannot be stopped because we cannot stop. There is nothing else; there are no options, there is no substitute, there is no limit to what we will do in pursuit of the goal. We simply can't. For all its downsides, this gives us a tremendous edge over people for whom music is something they'd merely like to do - however badly; or people who want to use music to be famous or rich. Those are also weak fuels to burn in comparison to the all-consuming, irrepressible need we have. It's a healthy expression in service of an unhealthy, permanent, damaged state. We need this because somewhere inside there is a hole which cannot be filled in any other way.

    Necessity is the mother of invention. This need provides for more creativity, more exploration, more expression than any algorithm can match, if for no other reason than it is about being human. That's what we're doing - that's what good music does - it connects, in a human way, with humans. It triggers experiences and thoughts and emotions, and when that expression comes from another human being it brings both sides to resonance in ways which defy explanation, but not observation. An AI will never suffer as a human suffers, and thus never fully have the resources to end that suffering through its art. It can replicate, it can extrapolate, it can invent, it can create. But ultimately, it cannot compete.

    It can and will replace tons of composers along the way for sure, but mostly for the same reasons it will replace humans in so many other ways - humans have gotten weak, passive, and stupid. With no natural predators we've become that which nature abhors. Our population continues to explode without limit, generating ever more flotsam and garbage, wandering aimlessly through short, pointless lives listlessly tapping out banalities on glowing rectangles with their opposable thumbs. AI is, hopefully, a force which brings this to a head. What will weak, soft, stupid people by the billions do, when they aren't even needed to alphabetize those insurance claim forms 40 hours a week? Or hold that sandwhich store sign on a corner, doing humiliating little dances in desperate hope of not having their "job" outsourced to a 2-by-4 and a nail? I don't know, but I'm looking forward to writing music about it. It'll be good, too. Not worried.
    Paul T McGraw likes this.
  9. I can appreciate that, but I meant most people aren't like you or Mike in that regard. For many their job is more about being needed somewhere, than themselves needing their job. Not talking about music specifically.

    I never did something like that (but I guess the question was rhethorical). Individuals are always able to profit in ways that aren't possible at a large scale. I don't see us having a "creative economy" where all artists live off vanity projects for others.

    THAT, is precisely what I'm most worried about. What will happen when the concept of "having a job" gradually becomes obsolete, we have no plan for that at all...
    Paul T McGraw likes this.
  10. That plan is a plan for population control and elimination of at least 50% of the current population. It's the only sustainable model and the only thing worth talking about. Of course nobody's talking about it.
  11. Thanos talks about it all the time.
  12. #12 Rohann van Rensburg, Mar 2, 2019
    Last edited: Mar 3, 2019

    This discussion comes up an awful lot and there are particular users on VI-C that love to jump on the dystopian bandwagon and speculate to death, rather than working on their craft.

    The philosophical arguments about the limitations of AI aside, a few things come to mind:
    1. I don't understand the tech obsession with trying to destroy humanity. Automation is already wreaking havoc on certain job industries, and the true underlying purpose is to feed company greed and maximize profit, not provide economic stability and further humanity's best. That said, the type of personality that facilitates tech development notoriously lacks the ability to grasp holistic philosophical thinking. There seems to be a pathological need for development at any and all costs.

    2.. We revel in novelty, but this is not where significance is found for us. Novelty doesn't last. When listening to this piece I think, "neat", and then immediately forget about it and stop caring, because there's no connection or human significance to it. It's not trying to "say" anything, because it's metaphysically impossible that it would have anything to say (the AI, I mean). Sort of like a generic pop song created as a radio hit. Catchy one day, forgotten the next.
    People care about what other people create, because it facilitates both a reflection of deeper metaphysical reality and the nature of shared human experience. AI can replicate that, but they can't compete, as Mike said. Doug raises good points about the fact that generic music is already available aplenty. I really have a hard time believing AI will replace art, because that illustrates a profound misunderstanding about what art does, what art reflects, and why art is significant. I mean record players and recorded music has been around for decades, but has it replaced live music? It will certainly replace a certain tier of composer though, at least in a commercial sense.
    But what does that mean, really? 50 years ago, you couldn't jam around on a synth and call yourself a composer. With the advent of tech and accessibility, the market is absolutely flooded with mediocrity. As misanthropic as it may sound, I don't think this is altogether a bad thing, the other implications aside.

    3. I know the feeling of despair. I really do get it. But then I remember: why am I doing this? Yes, I want to put food on the table for my family. I want to fulfill my basic duties as a father, husband and member of a community. But why did I get into music? To make money? Getting back into it at 23, I knew I couldn't choose a worse route for financial security. If you're in it for the gig, you'll probably be disappointed -- this is something I have heard literally every successful musician say when asked about being an aspiring musician.
    For me, and probably most others, it's a need. I have to make music. I have to create. It's unfortunate that we live in a time where creative personalities are not favoured (at least outside of silicon valley) -- North American culture does not value art the way it should be valued, and it's costing us eudaimonia, but it is what it is. That doesn't diminish the joy or sense of fulfillment I feel when making it, nor what it does for me spiritually/intellectually. It doesn't affect how it grows me and continues to further me into flourishing (in an Aristotelian sense).
    The upside of tech and globalization? I can hear my ideas played by a crappy orchestra, all at my computer. And as Mike talks about, there will always be an audience for your music. If you invest into your craft and your passion is clear, there will be someone who connects with it, someone to listen. I'd rather have two people who really connect with my music than have it heard by a thousand who don't.
  13. Great quote and interesting points raised all throughout this thread. AI is a topic I know a lot about as I implement it in my field of work (which is not music...)

    The main point to understand about AI is that it is simply is a set of algorithms to find and use organised patterns in large datasets. The science is a few decades old for the most part, but we are talking about it so much today because, as the western world is digitalising all human knowledge, there is a lot of data to feed these algos with and check the results.

    Now, If we assume that all theoretical issues around model convergence and scaling are solved - very big assumption - a computer should be able to model how a piece of music is structured from all the music scores ever written. Then, when you will ask a computer to finish the score you started, the result will be some sort of non-linear weighted average of all the patterns found in all the pieces ever written.

    For any innovation to truly happen, you would need to define a mathematical way of evaluating how "good" a piece of music is and the constraints (mood, tempo map, etc.) to be respected. In that case, the algos could solve for the most "optimal" piece of music and easily beat humans like it does in a growing number of fields.

    But as Mike points out, because artistic appreciation so subjective and so human centric, I am a lot less pessimistic about the impact of AI on artistic creation than on all other sectors. There is a case to be worried for composers who earn their living using a "set of magic recipes", but they do not bring much to the art form in any case, so no one will cry if we do not hear their music as much.

    From a practical standpoint, composers are already in strong competition to get commissioned, so machines do not change much from this perspective. The machines could eventually cover the box-standard requests, but when this happens, the public will probably grow tired of hearing it (too much of anything takes its value to 0) so the top human composers would still be in demand...
  14. Great points!

Share This Page