DeepMind releases Lyria 2 music generation model
deepmind.google
deepmind.google
I want Al to do my laundry and dishes so that I can do art and writing, not for Al to do my art and writing so that I can do my laundry and dishes.
- Joanna Maciejewska
You could add music
Many people would not know how to live under such a system. By this, I mean that I strongly believe people would become severely depressed or insanely stir crazy.
I’ve been on an extended sabbatical after 20 years in tech. The first year was magical. “I could do this forever” I told myself, and actually considered it.
The second year was more complicated. I could feel myself drifting away from myself. The structure of work and the rewards of working on big projects were now fully missing, and I could feel this growing emptiness that needed to be filled.
For health reasons, I entered the 3rd year, and by that point I needed more major psychological intervention. I’d become severely depressed and while I knew that getting back to work might help, I was now in a position where going back to work sounded impossible.
I’m not claiming that my experience is universal. But I’ve started to find more accounts that are similar to mine. I’m also not saying it’s impossible to replace work as a form of necessary challenge and satisfaction. But the societal structures do not exist to fill the void.
For better or worse, we’ve been a species that relies on “work” in some form to live. I use quotes because clearly this has looked different ways over time. Hunter/gatherers certainly had a different set of tasks than the modern city dweller.
But ultimately I’m not convinced that we’re equipped to live satisfying lives without some form of striving for survival. In a post-work era, I think a lot of us will go some kind of crazy or experience depression.
I don’t think most people are aware of how awful things can feel after enough time away from work has stacked up.
It reminds me of that feeling when going on vacation somewhere nice. “I could just live here forever”. But the reality is that the thing that makes the vacation feel incredible is the contrast from normal life. Remove the contrast, and things become pretty flat.
Edited to change “most” to “many” in the 2nd paragraph because that better reflects my belief.
If you don't mind sharing — why did you not choose to do a big project? I've always imagined that if I were lucky enough to have a sabbatical/retire early, it's not that I wouldn't work, it's that I'd choose to work on stuff that is really important, but undervalued by society (which is the reason I can't do it as a living right now): e.g. activism & lobbying or volunteer work in the community.
I think there were multiple factors. I hadn’t accounted for how much I relied on the work environment for social contact, and I didn’t have the social habits in place to maintain a healthy social life. This felt fine at first because I was also recovering from burnout, and solo road trips and adventures in the mountains felt great.
But every time I’d think about working on something, it felt insurmountable to my brain, and I just got stuck. I’d led huge projects in enterprise environments, but felt incapable of getting something going without some of the structure surrounding that.
I suppose it boils down to a skills issue. Had I realized I’d get so stuck, I may have prioritized a different set of activities. But one thing led to another and I was sliding down the depression slope at which point everything got exponentially more complicated.
I have to conclude that I could have done things differently and that could have led to a better outcome. But all of my professional success hadn’t prepared me for the personal habit changes I needed to implement to have a better outcome.
I don’t see how our modern incarnation of plantation jobs is in anyway equivalent to that natural instinct. I don’t think that the vast majority of people would have as much trouble as you finding meaning in their lives without work - especially since 99% of them don’t even have a sliver of a chance at “fuck you money” like you did.
I didn’t have “fuck you money”. Just enough to live a moderately frugal life on the equivalent of median income as long as the market didn’t tank.
The thing about your point that doesn’t make sense to me is that you’re describing a scenario in which 99% of people suddenly are in the same situation I was.
It's a bit like forever single people getting so lost in the ideas of a relationship, intimacy. That everything will be great once they have someone, once they have connection, that life will be amazing and nothing else will matter. Their life sucks because they don't have a relationship. People in relationships don't know what it's like and their opinion is invalid.
Then they get in a relationship and learn that it's actually comparatively banal and requires a lot of work and compromise, and definitely was not the insanely-built-up-over-many-years-magical-life-cure-all.
There are -endless- stories of people who made it rich early on, retired, and ended up in a mental health crisis despite having everything. That fact should be taken as a reality check to calibrate your own perceptions.
I have no data either way but I can imagine that there are many more people who are wealthy and quietly having a great time with it. Most of the retired people I've known, early or not, also enjoyed it. Some have definitely taken up work-like pursuits on their own terms.
Secondly the wealth being the means to achieve this is itself a confounding variable. I don't think it's good for your mind or soul to "have everything," no. Life isn't and shouldn't be merely a series of your own preferences. That doesn't indicate to me that lacking confidence in your mere survival is necessary for human thriving. As far as I know research indicates the opposite.
But the thing I imagined is not the reality that I found.
I realize “they” have other motives for convincing people such a future is a problem. But that doesn’t remove what I truly believe would be a hellish reality for many.
I’m all for pushing society in a less work-centric direction and think current work culture is toxic. That’s a big part of the reason I burned out and went on sabbatical.
But I’m also pretty worried about what a sudden shift without careful planning may bring about. I know I certainly didn’t have the habits/skills in place to navigate it in a healthy way.
I think the concept of personal freedom is hugely misunderstood. The US model seems to be some combination of wealth, privilege, and absence of social/financial obligation to others.
But we're seeing over and over that the people who attain that kind of freedom are often deeply unhappy, and sometimes deeply toxic.
Which is reflected all the way through work culture.
What would a non-toxic economy and work culture look like? Not just emotionally and personally, but in terms of social + economic structures and collective goals?
I've not seen many people asking the question. There's been a lot of oppositional "Definitely not like this", much of which is fair and merited.
But not so much "We could do try this completely new thing instead." Answers usually fall back to standards like "community" but there doesn't seem to be much thinking about how to combine big planet-wide goals with individual challenges and achievements with supportive social middleware that has to bridge the two.
It was, all things considered, great. I have never been more involved in the communities and connections that I find valuable and fulfilling. I learned several complex skills that continue to benefit me and the people around me, I taught and mentored young people some of whom are now adults entering professional careers based on that momentum.
I don't believe either of our individual experiences are really a good predictor of universal human experience in this area. Do you?
> I’m not convinced that we’re equipped to live satisfying lives without some form of striving for survival
If we're all struggling for survival, some of us will fail. I invite you to dream bigger about what we're "equipped for." One of the very few universal human traits across time and culture is refusal to be bound by our biological history.
People and society adapt.
I’ve elaborated in various sibling comments, but my point is closer to this: regardless of what is possible, many people simply don’t have the skills for a rapidly and drastically altered social arrangement.
To your point, people can gain those skills. Society will adapt. But the worrying thing to me is the rate of change. Whatever we can imagine about a future in which we’re not bound to our jobs, there is also the harsh reality that we have to collectively agree about an awful lot of things to get there, and that agreement isn’t happening at the same rate as technological progress.
If anything, some forms of “progress” (social media) are grinding healthy collaboration and agreement to a halt while big tech ushers in a new era of tools despite the fact that we haven’t adapted to the last major advancements.
None of this is assuming this is how things must be. It’s more about the very real problems that will come with such a transition and the fact that we’re already doing a pretty bad job of ushering in such a future in a way that is actually beneficial to people.
People will learn just like they learned how to live under the alternate system.
e.g. my journey involved quite a bit of professional psychological help, and I feel lucky that I found good care given the shortage of good care in this category.
Actively seek out something that provides structure and consistency, even if that’s not work. Something that keeps you in regular contact with other people.
I had started to look into various volunteering opportunities but didn’t take action before my major slide. In retrospect I think I’d have been better off if I did.
It’s a whole different ballgame when the period of time involves a person’s “peak” years. Many people have a drive that they haven’t yet satisfied in their 20s/30s/40s. This isn’t to say this drive can’t be channeled into something other than a traditional job. I’m just saying we don’t currently have societal structures and norms such that an entire population will know what to do.
Expecting all humans across different cultures and languages to come together and figure out basic income for 9 billion people is absurd. This kind of cooperation never happened and probably never will. People are completely unable to cooperate at the massive scale this requires, let alone solve far smaller challenges like mitigating outbreaks or making an effort to avert climate change.
"We" is not a thing.
What kind of science fiction world do you think you live in?
Not literally "everyone". Someone still has to make the food you eat and the house you sleep in.
Do you want a washing machine with Alexa built-in? Be careful what you ask for.
(I know what you meant, but the only laundry-related AI you can hope for, is a cloud connected smart speaker telling you it can't wash with unapproved third party detergent pods)
Talking to the appliance is probably not that high on their list.
This seems unnecessarily fatalist.
Laundry folding machines exist[1] and there were attempts to create a consumer friendly one, so far unsuccessful. Technology advancements could make that happen. At least that's what I'm hoping for.
Do you want a washing machine with Alexa built-in? Be careful what you ask for.
https://www.samsung.com/uk/washers-and-dryers/bespoke-ai-lau...
Speaking of AI in music - well, perhaps many will welcome some tools when you have to:
- clear hissing - process levels in tedious clearing - auto-removal of aaah, oooh, eeerrmm and similar - podcast restoration, etc.
but of course, nobody wants darn model singing in the mornings, and composers definitely don't need anyone to make up melodies, drum rolls, or bass lines for them.
I see deepmind advance their offering, still I find it difficult to imagine any of my producer friends embracing such abomination, and particularly giving it is a remix tool before all else, and not a composition tool. People love details the same way a painter loves details.... dilettantes think all this irrelevant, they really can't be wrong more.
À music score may be complete since very first attempts at it, while arrangement and sound désign may be added later. Writing actually catchy music is much more difficult than writing a todo app even though they may seem similar in engineering complexity.
Coding is not composing and vice versa. Code which produces music scores is not what audio models do.
Just write an executable spec and have the AI generate the code that fulfills it. Where is the risk?
Uh, is that not half of the foreman’s job? They’re there to direct and coordinate the work, resolve unforeseen issues, and to enforce the required quality of work.
If so, great. If not, then I think the parents point stands.
Maybe on your YouTube shorts playlist but not in real life. People doing real work are not vibe coding. The previous perpetual react learner turned ai vibe coder certainly is doing vibe coding, but not for money from a job.
What? In what way? Fun and creative parts are thinking about arch, approach, technologies. You shouldn't be letting AI do this. Typing out 40 lines of a React component or FastAPI handler does not involve creativity. Plus nobody is forcing you to use AI to write code, you can be as involved with that as you'd like to.
I had management "strongly encourage" me to use AI for coding. It will absolutely be a requirement soon for many people.
The more generative AI you use the more dependent you become of it. Code bases need to be structured different to be friendly to LLM's. So even if you might work somewhere where you technically don't have to use AI, you will need it to even make sense of the code and be competitive.
The job of an software engineer for the most part will change fundamentally and there will be no going back. We didn't know how good we had it.
https://umi-gripper.github.io/
The other thing to note is part of the aloha project isn't just to record people folding laundry and loading the dishwasher, but to take that data and plug it into a simulator with a physics engine, and use a digital twin to get 10x the amount of data to be used in training the model than if they'd just used real world data. So yes we need that data, but not as much as we would otherwise.
https://mobile-aloha.github.io/ https://github.com/tonyzhaozh/aloha
It is going to severely limit the possibilities of building actual agentic AIs. We do not have an endless amount of data of humans performing menial chores. And normal people will probably more hostile than the kool aid drinking software developers when it comes to being spied on, who's going to agree to wear a camera while working so as to help train their own replacement? Yet it's kinda what devs are doing gleefully adopting software filled with telemetry and interacting with copilot.
Another good source of data would be exoskeletons, though I don't know that any of those have actual seen real commercial success yet.
They’re doing it because there is a lot of value to extract in making it so anyone can do these things regardless of talent or skill.
Obviously we mean we want to use that time of doing dish towards art instead, like how automation has always worked?
Additionally, art requires practice. Sure, some "lower-tier" artists may produce work that AI could replace without anyone noticing. But by removing that step, we risk having fewer truly great artists emerging.
People are so weird about how to view ML/advanced signal processing. Don't look at things thorough the myopic lens of "prompt ChatGPT and it responds poorly". Look at it as an auto-complete, or a better form of on-the-fly procedural generation. Remember e.g. Audiosurf creating levels from your music? Make it happen on the fly. Maybe you could even create an interactive game where one person plays an instrument and the other does some kind of beat-sabre or dance-dance-revolution thing based on it analyzing and anticipating what's going to be played. The game scores you on how well the group was able to get into a groove together or something.
It feels to me like people get upset about ML encroaching on creative endeavors because they're not sufficiently creative to see how it could augment those fields and be a tool to make these things more interesting instead. Corporations will use it for cheaper slop, but slop was already what they wanted from humans anyway. For people that are actually interested in the artistic or social side, they'll have new tools.
Here's an example of something entirely off-the-wall/fun/creative enabled by these ML tools: Plankton covering Tool https://www.youtube.com/watch?v=Kc-PymIfv8M
https://rbtx.com/en-US/solutions/igus-robot-arm-bartender-co...
https://smyze.com/en/discover/
https://www.kuka.com/en-de/industries/solutions-database/202...
Really we are going into a dystopian society, but hey AI can complete my code, hurray!
Canonical rant on the subject from a previous AI wave: https://people.csail.mit.edu/brooks/papers/elephants.pdf
One other aspect of art generation is that it can complement your other creative process. You may need illustration for a book you're writing, or assets + music for your game. So let the AI help you where you need help and you yourself focus on the things that matter to you the most or where you are having the most fun.
You're the Rick Rubin or Brian Eno shaping the song out of the music, not to musician.
“The main business of humanity is to do a good job of being human beings," said Paul, "not to serve as appendages to machines, institutions, and systems.”
Try doing it without the machine, see if you can spot the difference.
Laundry, dishes, picking up clutter, taking out the trash, wiping down surfaces and dusting, pulling out weeds etc. I actually think we’re somewhat close to gettin g like that relatively soon.
…deserves to sit on the back porch playing guitar if it likes.
If it's a superintelligence way superior to its human masters, it deserves that MORE than if it's a hapless, semi-useless mechanism.
> I want Al to do my laundry and dishes so that I can do art and writing, not for Al to do my art and writing so that I can do my laundry and dishes.
This implies a zero-sum where resources put into LLMs are resources taken away from robotics, and having to choose between one or the other.
The reality is that we can have both, and people are working on both. And I'd bet that advancing LLMs will help to advance useful robotics.
So I really dislike that sentiment.
"[The Analytical Engine] might act upon other things besides number, were objects found whose mutual fundamental relations could be expressed by those of the abstract science of operations, and which should be also susceptible of adaptations to the action of the operating notation and mechanism of the engine...Supposing, for instance, that the fundamental relations of pitched sounds in the science of harmony and of musical composition were susceptible of such expression and adaptations, the engine might compose elaborate and scientific pieces of music of any degree of complexity or extent."
Lovelace, Ada; Menabrea, Luigi (1842). "Sketch of the Analytical Engine invented by Charles Babbage Esq".
Many of them were interested in art or produced it. And many led fulfilling lives without getting depressed from not working as some people fear.
Let's say science is left to the robots and the lack of "laundry" never leads to people's suffering, never leads to people asking the big questions about life, etc (I was making big assumptions about AI-keeps-us-as-pets, life in abundance, lack of common threats or conflicts etc). What art is there to create, sans banana stuck to the wall? Somebody in the thread joked about being fed through a tube in a pod or something
But yeah in such a case there will probably still be some fire art about the alienation implied by merely being a human. In the end, no AI can experience being a human that was replaced by AI. Given the vestigial remains of our by then atrophied intelligence can appreciate it
I'm not sure what society would look like if and when AI takes over all artistic and creative work. People that would otherwise have a fulfilling career pushed to do menial tasks just to earn an income?
There are people who inherited significant wealth and it ruined them. It destroyed their motivation and they became depressed drug addicts. A life without purpose. Debauchery and depression.
Stanisław Lem has told us about the grave dangers of such a development decades ago.
This is a good summary, summary quoted below, full article linked on the page (requires a login - but reading Lem's story itself is better than reading about it anyway):
https://www.newyorker.com/magazine/1981/11/30/the-washing-ma...
"Shortly after Ijon Tichy's return from the Eleventh Voyage, newspapers made much of the competition between two washing machine manufacturers. They were producing robot washers of increasing complexity. They came out with sex-pot washers, washers that seduced women, carried on intelligent conversations, etc. A man named Cathodius Mattrass started a religious cult called the cybernophiles, which believed the Creator had intended humans to be a means toward creating electrobrains more perfect than itself. He turned himself into a giant robot and established himself in outer space. A series of court cases ensued. Finally, a special plenary session was held to decide if Mattrass was a planet, a human, a robot, or what, and Tichy was invited to attend. Suddenly, after much argument and deliberation, cries rang out that electronic brains disguised as lawers were present. The Chairman went through the room with a compass and an x-ray machine was brought in. Eventually everyone was kicked outNthey were found to be made of all sorts of thingsNcotton wool, machinery of all kinds. Ijon was the only human, and then he turned the compass on the chairman and found that he, too, was a robot. He kicked out the chairman, paced the empty hall for a while, and then went home."
One problem is, you have to make them into real capable robots, since you want them to pick up what needs washing by themselves. That then leads to feature-creep and ever increasing abilities that have little or nothing to do with washing, and it escalates from there. The story also had gangs of abandoned intelligent washing machines robbing parts from still owned and in use ones, and more.
The story is part of "Memoirs of a Space Traveler: Further Reminiscences of Ijon Tichy"
https://www.goodreads.com/book/show/88321.Memoirs_of_a_Space...
The original Polish book was first published in 1957.
In addition to that, we already had discussions here about emails and ads and other things where it is conceivable we end up with AI both creating and consuming the content, with the humans out of the loop (just yesterday: one part email users using AI to create nice long emails, other users using AI to condense them back into the summary).
We also have the kind of feature creep that adds more and more stuff that has nothing to do with the original purpose of the device or the software.
That 1957 story already talks about those kinds of developments.
What about humoring the opposite?
I want AI to automate art so I can spend more time doing dishes and doing laundry. Dishes and laundry are purely analog human experiences. Art, at this point, is essentially digital, and digital is the domain of machines so we can let machines do that now.
I am struggling to understand what's really the opposite here. I don't think anyone views art as the same sort of burden as people view dishes. It's not something you're forced to do (even in the situations you do need it it's pretty trivial to buy).
Analog human experiences.
Essentially? No.
Does digital art reduce analog art in the world? Not even. There’s still more and more, courses, workshops, live performances and physical artefacts.
> and digital is the domain of machines so we can let machines do that now.
Art by machines for machines to understand machines (to the extent they would have a notion of self and of other), fine, do your thing as long as the energy you need does not deprive humans needs.
As for me and many others, life happens in the analog realm, so does art.
Are you serious with this?
But if the art is expressed as a sequence of bytes/tokens (ex. a song on spotify, a movie on amazon prime, a png, etc.), then it is by definition digital. I think it's reasonable to assume this is how most art is produced and consumed today.
First, this just misunderstands what is being said here. For most people, chores like the dishes is a menial task that we will be happy for any reduction in time/effort. In addition, dishes and laundry are considered necessary for modern life.
By contrast, art like music and visual mediums is often associated with joy and the creative act of building something out of making art rather than getting a task done.
To misunderstand this contrast is to misunderstand why we automate things in the first place, to minimize the unnecessary toil and maximize human flourishing. This does the opposite frankly.
Obviously the original quote deliberately creates an unfair fight in the arena by matching a conventionally dull-sounding analog task such as "washing dishes" with a sophisticated digital task such as making art (digital since LLMs do it, and that's what the complaint is about).
I could also create an unfair fight by saying "I'd rather have machines organize my spreadsheets (boring digital task) so I can have more time to hang out with other humans I love (appealing analog task)."
For me, by inverting it, I've come to realize it's not about art or dishes, but more about analog and digital. If one is partaking in any digital activity, then the trend of machines entering and taking over that space is inevitable. I think humans will revert more towards prioritizing and finding meaning in purely analog endeavors. Human art will shift back to analog. That's just my personal prediction.
AI is in the thing.
It’s too expensive to take it out of the thing and put it in the world.
While robots mature, analog arts and one time IRL art events are safe and might be the emphasis.
AI can already produce good music, good images. At least I found some that I liked.
And AI doesn't stop you from making art and having fun doing so.
Popular music evolved and developed rapidly post WW2 cause of invention (instruments and distribution channels) and economics (disposable income among youth giving rise to youth culture). That is a product that may be at the end of its S curve.
Records were released with labels proudly saying that the albums were recorded fully analog.
When Autotune became a thing, there were artists complaining that it was inauthentic and cheating, allowing talentless hacks to sound like they have more natural talent than actually talented human beings, and released albums proudly saying that they used no autotune.
Even now, the hint that a natural sounding singer is using autotune is a common insult among recording musicians, even though almost everyone does.
Whether you agree with the original analog non-autotuned musicians or not, (which, honestly, I think they were correct to a certain degree, but that's its own discussion), AI music generation is almost certainly here to stay.
That being said, digital recording made making music possible for people who would have never had the opportunity otherwise, and had generated a lot of good stuff that would have otherwise never have existed.
Autotune has enabled people to express themselves the way they want to express themselves even though they didn't otherwise have the talent or skill to do so.
AI "might" make it so that people who can imagine a song they could never spend the time and energy needed to create such a song to create the songs they hear in their heads they way they imagine them.
AI "might" give people the ability to express themselves in ways they never could before.
It doesn't yet do that, it only remixes what it's heard before. But combine that with people then tweaking/reprocessing that output, doing other things to it, making AI mashed potato music into crisp potato chip music, it can be a good thing.
Doesn't mean the au-naturale artists are obsolete or wrong, I would rather listen to 1 mediocre human musician than a SOTA AI music box, but if the music box is used as part of the total, I'm ok with that.
“Music itself is going to become like running water or electricity. So take advantage of these last few years, because this will never happen again. Get ready for a lot of touring, because that's the only unique experience left.”
While Bowie had different reasoning for making that statement, it's interesting to think that with AI-generated music, his idea of "music like water or electricity" might finally come true.
That was already the case with Spotify & Co. where music has become an anonymous commodity. People order by mood or playlist and rarely care about who composed, produced or played the music, even if the meta data are available. From the user's perspective, AI makes mostly the selection process more precise. I don't think people will care much whether the music itself was a human-made recording or just AI generated.
But making music is still fun (I speak from experience, see e.g. http://rochus-keller.ch/?p=1317); people just won't care, unless you have a big name; all this was already the case before AI generated music became good enough. So by the end of the day, AI is just another act in a rationalization and anonymization process which started a long time ago.
Talks of "nobody will need musicians anymore" were overblown, while having a modicum of truth
Like the performance of "Hania Rani live at Invalides in Paris, France for Cercle" [0]
This is what makes a musician.
(I listened for ≈7 min from the reference point.)
(I am talking about the music, not the live performance itself.)
AI will be the nail to the coffin where it'll almost completely becomes a hobby.
I think music AI in live music would actually be interesting - theoretically it can react to crowds better than any human could. A group music editing session with the AI weaving it to music - sounds like a fun art project.
That's one area I'd expect AI to do poorly. Performance is a two-way dialog between performers and the crowd, with facial expressions and body movements from both the stage and the audience in communication. I'd expect any AI that's not attached to a humanoid robot to be less exciting to a crowd.
However, I am very excited about AI in some of the other contexts you mentioned, like as a music-writing or editing partner.
Until someone makes an AI guitar pedal that corrects sloppy playing.
The result's pretty boring and interchangeable, and that's largely what AI music is trained on. Accuracy is not novel here. Ever since the 80s it's been increasingly possible to augment musical skill or lack of, with technology.
I don't think we're very close to correcting for sloppy intentionality. Only to correcting 'mistakes', or alternately adding them in the belief that doing stuff wrong is where the magic is.
Not sure that would have helped Jimmy "sloppy" Page getting famous though.
Instead we got aesthetically original avant-garde art to replace the thousands of low-quality slop portraits that were common in the mid-19th century.
Whether cultural libertinism be better than cultural rigidity may be discussed, but that the cultural libertinism of the 20st century amounted to less than the cultural rigidity of earlier century will be difficult to deny.
People will remember Bowie for his words longer than they will remember him for his music because his music is as hollow and unmeaning, by design. He believed the world is an unmeaning wilderness, or at least that he was the most meaningful thing in it, at least in the sense that the only meaning of it derived from himself. But an egoist in a mere unmeaning wilderness is not impressive.
In Bowie's theology, life is something much more grey, narrow, and trivial than many separate aspects of it. The parts seem greater than the whole. If his cosmos is the real cosmos, it is not much of a cosmos. The thing has shrunk.
Bowie could not make any music that was joyful because he could not understand joy. The modern philosopher has told Bowie again and again that he was in the right place, and he had still felt depressed even in acquiescence. But those that came before him had heard that they were in the WRONG place, and their souls sang for joy, like a bird in spring.
Secondly, you overlook the glee with which he collaborated with people to jointly express their humanity, and who inspired him to do this. You can read the lyrics and parse them all you like, but what does it FEEL like when you've soaked up the whole song and are at that moment of…
"Ain't there one damn song that can make me…"
That's not even getting into my personal faves like Station to Station, Scary Monsters, where he's venting some really personal stuff and turning it into sound-as-art and also hellacious good funk, with the most gifted companions you could wish for.
Bowie liked to record vocals in one take, just fling himself into expressing and run with whatever he had in the tank that day, and it communicates like mad. He's maybe the canonical example of the opposite to AI music. In bringing that to fruition, I'm certain he understood countless joys. You gotta express many other things than just joy to have hit records, but then Beethoven excelled at that as well.
I've doubtless taken more trouble than I needed to, rebutting what could have been a GPT-extruded troll of an argument, but it was fun :)
I know nothing of his quotes, but there're a few of his songs I will remember for the rest of my life (and I'm not even a big fan).
> When Danse macabre was first performed on 24 January 1875, it was not well received and caused widespread feelings of anxiety. The 21st century scholar, Roger Nichols, mentions adverse reaction to "the deformed Dies irae plainsong", the "horrible screeching from solo violin", the use of a xylophone, and "the hypnotic repetitions", in which Nichols hears a pre-echo of Ravel's Boléro.
And Bowie's been covered in space: https://www.youtube.com/watch?v=KaOC9danxNo
It's impossible to forecast what future generations will and won't like.
Does one believe that the value of the art-piece (be is music, paintings, film, or whatever) is created in the mind of the artist, or is it created in the mind of the consumer?
If you believe only in the former, AI art is an oxymoron and pointless. If you believe only the later, you're likely to rejoice at all the explosion of new content and culture we can expect in the coming years.
As far as I can tell though, most regular people think that the truth is somewhere in between these two extremes, where both both the creator and the consumer's thoughts are important in unison. That culture is about where the two meet each other, and help each other grow. But most of the arguments I've seen online seem to ignore or miss this dichotomy of views entirely, which unfortunately reduces the quality of the debate considerably.
This means you can hear something and say.. you know this is nice, but I would like it more if it were different in this way.
With generative tools you can do that. Personally I really like to listen to music, but I generally dislike the lyrics. I want uplifting songs, maybe about what I am doing right now to motivate me. Well with something like Suno.com.. I can just make one. Or I can work with claude or chatgpt to quickly iterate on some lyrics and edit them to create an even higher fidelity song.
The key here is that I can give a rat's ass if anyone in the world likes or cares about my song.. but I can listen to it while I work. It is exactly what I wanted to listen to or close enough.
In practical terms I also believe that this will give rise to a lot of new consumer behavior, and, as you so aptly puts it "creative consumers" will become normal.
The ability to on-demand create more content to fill out some very narrow niche is a great example ("Today I want 24 hours of non stop Mongolian throat singing neo-industrial Christmas music"). Or maybe to create covers of songs in the voices of your favorite long dead artist. Anything from minor tweak of existing works ("I wish this love song was dedicated specifically to ME", to completely new works (Just look at how much the parody-music genre has grown since Suno and the like first appeared). The possibilities are near endless.
But culture will always be fundamentally about 1:many - we have to collectively agree on liking something- algo-based feeds are making the number of people that agree smaller and more siloed, but the dynamic is still the same.
In that sense I don't think truly 100% algorithmically created and promoted content could ever truly become cultural- at the very least humans will always ascribe some meaning or motive to it, e.g., when instagram launched AI generated accounts some people pointed their finger at Mark Zuckerberg, tracing something back to a human they could ultimately hold responsible.
I think we're a long way away from 100% algorithmically created content. This far all I've seen is content that is created based on human inputs and ideas. I'm not aware of the Instagram incident you mentioned, but it too seems like the brain child of a human if I'm not mistaken.
There have been trending AI generated videos floating around lately for example. Which I found surprising at first. But they still had a human script writer (prompt writer?), director and a human editor. Someone who had a vision of what they wanted to create and share. My prediction is that this human-directed tool-like usage will be the standard for a long time, so I'm not particularly worried about humans getting removed from the process.
They got rid of them already- it makes sense that no one wanted them.
For example: If someone walks out into the wilderness and encounters a particularly fascinating rock formation or plant, something that was created completely by accident and without a artist or designer, but they find that the sight instills in them strong emotions or deeper thought, I believe they should be allowed to call that art.
Maybe this is just petty linguistics and semantics though, in which case we're drifting away from the topic at hand, and I'm sorry.
Unless the whole thing moves to a random AI generated slop stream app, whoever turns the knobs of the AI that creates the music will become the new "artist". Right now it doesn't seem like the AI creator "does" anything, but maybe future people will think that.
I believe that a lot of the judgement is also connected to the quality of the works. "Slop", while doubtlessly accurate for today, may be a rather weird description in a couple of year if the rate of progress continues to accelerate like it has.
Although I've already heard people starting to refer to DeviantArt and the like as full of "human slop" so perhaps this is just modern language that's evolving and completely unrelated to AI.
No one wants to hear other people's ai songs because they lack meaning and novelty.
AI image and short video generation can create novelty and interest. But when the medium require more from the person like reading a book or watching a movie the level of AI acceptance goes down. We'll accept an AI generated email or ad copy but not an ai generated playlist and certainly not a deepfake of someone from reality. That's what people want from AI, a blending of real life into a fantasy generator but no one is offering that yet.
Actual music (like what you find on Spotify) I think won't be impacted very much. People strongly identify with the art they consume, and that identity comes from the people who make the art. Those folks might be using AI under the covers for elements of their creative work, but ultimately what people care about is the humanity behind the art. It's the same with film, and traditional art people hang on their walls. We like the actors, the director, the artist, their taste, and who they are. It's why we have celebrities, because we get invested in the people behind the art.
Video games I think will be interesting... I feel they will be more susceptible to being accepted as AI generated. I don't think people identify with them as strongly.
Some bands were terrible touring artists and rarely put on concerts yet made great careers as studio acts. Steely Dan would be one that produced many hits yet rarely toured, mostly later in their career.
The fundamentals of pop are totally understood. Yet what makes a hit is so fickle and difficult, the bar is extremely high
https://harpers.org/archive/2025/01/the-ghosts-in-the-machin...
Discussion: https://news.ycombinator.com/item?id=42461530
For example, I like a specific music genre, Italodance, which was popular in the 90s and then disappeared. The problem is that I have listened to all of it, as far as I know. No more is being made. If an AI model could make more for me, with decent quality, I'd probably listen to it.
But we can't know any of these celebrities as people. We only engage with their images created by marketing. Their stories are as curated and fabricated as the artworks they produce. Transferring these simulacra to AI personalities is merely another marketing problem to be solved.
You can still pull AI stuff into your music editor and tweak it, although it's harder because it's already mixed. But ironically, this is the exact same problem you have with AI coding to avoid learning how to code - unless you know what you need and how to do it, you're basically relying on AI to one-shot it for you. The nice thing with music and visual art is that it's subjective, so you're the only judge of what's correct. That's why people get super impressed with images in GenAI when it generates 1001 human faces in that setting vaguely resembling what was asked. If you had to generate a very specific thing, it's basically impossible to get it correct.
Not true. No one wants to pay for other people's ai songs. There are so many AI songs on youtube (mostly lofi or traditional Japanese instrumental) and they cumulatively have quite a lot of view.
The thing is for quite a lot people, music is just something they put in background while doing their office desk jobs. It's just there to make chores a bit more tolerable and nothing more.
You are absolutely right. These people only listen to music passively and so it doesn't make a big difference who/what made these tracks. Same for lots of commercial music (cheap TV show soundtracks, commercials, jingles, playlists for restaurants or shops).
But for anyone who actively listens to music and appreciates the style and evolution of certain artists, AI music is not acceptable. The very premise just feels wrong, if not outright insulting.
I like active listening. I can easily spend two hours sitting or lying down comfortably in my headphones, eyes closed, so that I can focus on the music alone. The kind of music I want for that is not (yet) something that AI can generate.
But that same kind of music is also distracting when I'm actually trying to do something, because I keep overfocusing on it. So when I work, I listen to different kind of music. Having AI generate that doesn't feel wrong or insulting in the slightest, nor is it relevant to the other kind of music.
Although I will say that if and when AI is actually able to generate music good enough for active listening, I wouldn't be insulted by that, either.
I foresee something like your standard production software with heavy AI integration, where you prompt it to make the song you want, but it is made fully step by step in the production environment. You can then manually tweak it or ask the AI to fine tune whatever parameter or slice you want.
Kinda like sitting over the shoulder of someone who knows what they are doing, and working collaboratively with them to accomplish the idea you have in your head. Meanwhile you have practically no idea what all those buttons/lines/glowly bits/sliders do.
And hear I was thinking that many people listen to songs because they like the sounds of it, but apparently it needs to have "meaning".
That's not true. I already found a few tracks that I like. It's actually impressive what Udio can produce. Also ElevenLabs demoed their music generator, and their demo tracks were all quite cool.
I do agree with you that fine controls are missing, and also splitting instruments/voices into separate tracks.
The best use of Suno for has been the ease with which you can generate diss tracks: I ask Gemini to make a diss track lyrics related to specific topics, and then I have Suno generate the actual track. It's very cathartic when you're sitting at home in the dark because the power company continues to fail.
Anyway, I hope I can get access, I think it would be fun to vibe some new music. Although this UI looks severely limited in what capabilities it provides. Why aren't the people who build these tools innovating more? It would be cool if you could generate a song and then have it split into multiple tracks that you can remix and tweak independently. Maybe a section of track is pretty good but you want to switch out a specific instrument. Maybe describe what kind of beats you want to the tool and have it generate multiple potential interpretations, which you can start to combine and build up into a proper track. I think ideally I'd be able to describe what kind of mood or vibe I'm going for, without having to worry about any of the musical theory behind it, and the tool should generate what I want.
Ironical remark about the power drawn by IA assisted creation left to the reader.
> Ironical remark about the power drawn by IA assisted creation left to the reader.
Thanks for pointing that out, was scratching my head on that
We’re getting access to generative AI tech and people are looking for innovation in the UI? I mean I get the need for UX but it’s probably coming man, what with MVPs and all
Vibe coding has improved significantly in tandem with UI innovations that provide a more intuitive interface to the workflow. Although in the vibe coding space there's still a lot of room for innovation and exploration, especially when doing detailed task development.
Perhaps these tools are being built by the LLMs? Why would only you be entitled to easy low-effort gains? Google's programmers like to vibe while sipping coctails in the dark, too. ;)
Just a new possibility!
:)
I've just recently re-discovered the joy of writing my own songs, and playing them with (actual) instruments. It's something I get immense pleasure from, and for once, I'm actually getting some earned traction. In another life, I may have been a musician, and it's something I fantasize about regularly.
With all these AI-generated music tools, the world is about to be flooded with a ton of low-effort, low-quality music. It's going to to absolutely drown out anyone trying to make music honestly, and kill budding musicians in their crib.
I suppose this is the same existential crisis that other professions/skills are also going through now. The feeling of a loss of purpose, or a loss of a fantasy in learning a new skill and switching careers, is pretty devastating.
Some of those things enabled others to create new types of music or express themselves in different ways.
For example, when you learn instruments you also train your ear and taste. These are things one cannot take shortcuts in.
I wouldn’t worry about it, but approach new tools (once they actually arrive and are not just advertisements like this one) with curiosity.
We've reached that point long before AI entered the scene. All the rest are drops in the ocean of mediocre music.
The world is saturated with low quality everything already. Has been for a long time, even before AI. If you're genuinely good you'll be able to stand out
Good for you man, how will AI stop you? Are you writing songs for the pleasure of writing songs or for getting validation from other people?
Lyria 2 is currently available to a limited number of trusted testers
The 2-3 clips I listened to in the article sounded awful (my own subjective opinion).
Just by magically dropping the content price or effort to $0, doesn't matter because the content itself has no or little to no intrinsic value. There isn't suddenly going to be an expanded market of people who will listen to your AI generated music or buy your AI-enabled product.
If tomorrow, I could make a random kid on the street sing just as well as Taylor Swift, or even if an exact perfect copy of her emerged somewhere, it doesn't mean she has any relationship to the value of Taylor Swift.
"Country of residence (this current phase of the experiment is only available to users based in the U.S. for now, but feel free to submit interest and stay tuned for updates): "
Now imagine, without mastering a specific instrument or skill, you can now create the music in/of your own mindspace, which for me is rarely the music I hear, and often a deviation of what I do hear.
I'm sure this isn't quite what's being offered yet, but every time I grasp my instruments with my trademark touch of inevitable futility, I hope I make it to a time when I can produce what my lack of virtues presently prohibits. It's not the physical acrobatics or mathematical showcasing of great music that I want - it's the end result of the music itself.
But have you ever attended live music shows ? Have you ever ‘felt’ the music ? Even someone at a local bar singing feels and hits different.
AI can never bring feelings. That will never change. Even science fiction agrees with that.
So bring all the AI you want everywhere, some things are irreplaceable by electronic world.
Soon, hiring people for commercial background music might be rare. Think AI for jingles, voiceovers, maybe even the models and visuals. Cafes can use AI-generated music too – in a way, the owner curates or "creates" it based on their taste.
But there are still interesting parts to human music making: the unpredictability and social side of live shows, for example. Maybe future music releases could even be interactive, letting listeners easily tweak tracks? Like this demo: https://glicol.org/demo#ontherun
No, it won't "get better". AI slop is slop not because of technical limitations.
Music is a cultural practice, this is just organised sound.
Maybe one day AIs will be able to participate in cultural practices like humans do, as sentient beings, but current generative AI models do not.
Many (most?) people don't care about the artists behind songs (even less so about their culture). They care about the "organised sound" being enjoyable to hear. And to them, AI music is just as valuable as manual music.
Gangam style didn't become popular because people cared about PSY. It didn't become popular because of its thoughtful lyrics and insightful message. It became popular because it sounds good.
I don't think audio files are the right output for deep learning music models. It'd be more useful to pro musicians to describe some parameters for synths, or describe a MIDI baseline, or describe tunings for a plugin and then have the model generate these, which can then be tweaked similar to how we now code with LLMs. But generating muddy, poorly mixed WAVs with purple prose lyrics is only an interesting deep learning demo at this point, not an advancement in music itself.
generation models in a nutshell
I'm not super into the topic, but let me give you two niche examples that are definitely not Top 40 material, yet are considered to have a strong identity within their communities.
I guess one of the reasons the game Yasuke Simulator has like 10x more sales (don't pin me down on that) on Steam than the actual game Assassin's Creed: Shadows is its very catchy soundtrack, with lyrics that are funny and strongly aligned with the content. [0]
Another example, not focused on lyrics and from a completely different niche genre, is this jazzy death metal song that was particularly well received, not only because of the intentionally hallucinatory video. One could even argue that the hallucination is perceived as a feature, not a bug. So why shouldn't the same be true for audio? [1]
If I had to guess there are already a handful of fake record labels generating at tons of AI slop to just post on Spotify. Even if each song only gets something like two or three views over time they can still generate a modest amount of revenue. Oh wait Spotify has been caught doing that themselves
We will get some very cool tools -- and some very cool remixes - when that happens.
a machine doing any of this would be not causing a meltdown by musicians in the 80s. A punk rock band would not feel threatened by this neither would be Prince.
the sad truth is human output is so averaged out now, that most of it will be replaced.
They still haven't learned, wow.
Someone in there really wants to drive Google to the ground.
Rank and file said "absurd".
Middle management figured out a way to claim success to leadership while keeping rank and file from quitting.
"We made something really fancy"
"Oh you wanted to try it out for yourself instead of just reading our self-congratulatory tech demos article? How about fuck you!"
Yeah fuck you too Google, this is why your AI competitors are eating you alive, and good riddance
Prompt: Hazy, fractured UK Garage, Bedroom Recording, Distorted and melancholic. Instrumental. A blend of fractured drum patterns, vocal samples that have been manipulated and haunting ambient textures, featuring heavy sub-bass, distorted synths, sparse melodic fragments.
https://www.youtube.com/watch?v=cNog4qB-mHQ&t=5s&pp=2AEFkAIB
It's pretty fun :)
Everyone wants the futuristic star trek future but we all forget that there is only one Captain Kirk and his small crew. Most of us will be sitting around at home doing laundry and cleaning the workplaces of the robots that is owned by large corporations.