Suni ai and Udio...the end of us?
- Thread starter pulpfiction
- Start date
jbuhler
Senior Member
Do you want to make music or do you want to prompt it?
It may be that someone figures out how to make a virtual instrument using these technologies. Maybe that happens this summer, maybe in a year, maybe in five years, maybe ten years… or maybe generative AI is a bust in this domain and it never works properly.
So I don’t think you can wait to buy if you need instruments. But neither should you buy now if you don’t need instruments. And if you are a beginner use free instruments or buy some inexpensive ones to learn with. Any skills you learn won’t go to waste. Meanwhile the time you aren’t making music is time that you won’t get back.
Crevalation
Active Member
Understand where you are coming from. it is worth noting this is not transactional AI to use for single purpose stuff such as audio restoration, Mastering (eg Ozone), increasing pixels etc. It is a generative, constantly feeding on new audio material, not limited by human constraint (only by computing resources), and improving at exponential speeds. It is also not like sample library where its “training” is constrained by human, needs to record/rerecord samples，massage samples, iterate by hand etc (just ask Performance Samples how long it takes for Pacific Strings Solo).
We are only at version 1, no version -1 actually. And it will be exponential, like so many progress of so many generative AI we have seen. And, there might be also other better competing technology/algorithm being born in the future, based on the past experience.
On the positive side, (not at the current version yet) - but when it’s more tweakable and flexible in therms setting precise timing and other parameters, I can see media content creator, Director and composers themselves using this to iterate ideas, just like how we iterate and brainstorm ideas using Dalle. Also, once it’s mature, I won’t be surprise if (already) there’s a commercial version catering to this need, where the timing and practicable parameters are tweakable to suit movie/video production needs. Will it replace composers, probably not, because there’s still the working relationship, experience, ideas and attempt to break new grounds. And also budget. I guess is, depending on the types of projects, in the near future (with an exponentially improved version), this will increase the expectation of the what a composer can do and the time taken to produce the goods.
MarcusD
Tinkerer
This is very true. Hard work is rewarding.
A lot of people are getting disheartened, even friends talking about giving up pursuing music. But, at the same time - if we just give up on creative pursuits (because of what this tech can do) it will allow such tech to become normalised a hell of a lot faster, with little to no regulation. Keep creating!
I absolutely agree with you on this, it will get better, the fuzzy sound is going to be less of a problem as models are trained on more data and algorithms improve. I am not arguing the fact that this will cause ripples on multiple levels, but at the same time I feel that given the capabilities of AI right now, people tend to sense impending doom where there is none.
Currently I would argue that text-based AIs need the least amount of time and resources to train. As a result, just to bring up an example, there are lots of people who publish their AI-written books. I've read recently that a year ago Amazon was flooded by AI-written books, this number has increased massively by now.
Does this mean that it would be harder for a writer to sell well? In a way, yes, but at the same time the human element, namely originality and creativity is missing.
Models could contain hundreds, thousands, millions or more different patterns of emotions. But that does not mean that those emotions are evoked on the same level for everyone. In other words, these AI are still like a sample library, snapshots, just on a much bigger scale, even if they are evolving. Weighted, so that the next probable block could be added to the end result, something that has been done numerous times before.
This means that for genres where there is an obvious pattern, such as with trailer music, an AI will be a lot more successful in attempting to create a track. But this only goes so far.
On the other hand, I can absolutely imagine AI-assisted composing even with the models we have today. Whether this is cheating or not I'll leave for another topic though. What matters most though is music. The art of creation. And I don't know about you, but I don't see myself giving that up even if AI can do ten times better compositions than all the classical composers together.
AnhrithmonGelasma
Senior Member
"Swift has said time and again that making music is simply just something she loves to do.
“I feel this happy when I finish a song or when I crack the code to a bridge that I love,” Swift said ... while accepting the Grammy for album of the year (her fourth). “For me, the award is the work.”"
"No, seriously, Taylor Swift could become an even bigger deal after ‘The Tortured Poets Department’"
RickOli
New Member
I think it's almost impossible for UDIO to stop progressing now that it has shown superiority, especially with powerful backers financing it. The way forward is for us to adapt, just as artists have adapted to tools like Midjourney, harnessing creativity. The next industry to be affected will likely be the film industry. With the advent of SORA AI and others, soon every corner will have a fantastic short film scriptwriter. In the future, amateur directors will create spectacular films using only AI commands. Those with a consolidated audience, followers, etc., will not feel the impact as much, because the key is to adapt.
I believe that there will be significant updates in UDIO for professional musicians in the future. These updates will allow them to manipulate instruments, create melodies theoretically, and convert each instrument's sound into MIDI, thus enabling them to modify and create something original.
MarcusD
Tinkerer
I don't mean to sound rude, but I really don't want to read an article about Taylor Swift. Can you summarise the point for me?
Trash Panda
Clueless nitwit
I’m sure ChatGPT could!
Certainly! Here's a bullet point summary of the article about Taylor Swift's upcoming album release:
- **Anticipation for New Album:** Taylor Swift is set to release her new album, "The Tortured Poet's Department," which is highly anticipated by the music industry.
- **Impact on Music Sales:** Swift's releases typically generate significant sales, including streaming and physical copies, disrupting standard sales patterns and chart placements.
- **Marketing Strategy:** She effectively uses merchandise bundles and exclusive editions to enhance album sales.
- **Industry Influence:** Swift's release strategies significantly influence music industry practices, affecting other artists' chart performances due to her dominant presence.
- **Control Over Music:** Swift maintains a tight grip on her creative output and public image, reinforcing her role as a pivotal figure in the music sector.
- **Overall Significance:** The release of "The Tortured Poet's Department" is set to reconfirm Taylor Swift's status as a major influencer in the music industry through her artistic and strategic choices.
David Cuny
Grand Poobah, Royal Order of WordBuilders
UDIO has shown superiority to what? Composers? Musicians?
Or perhaps ROI for powerful backers?
People who use Midjourney are real, bona fide artists, not just prompt monkeys. Good to know.
And prior to the existence of Midjourney, artists were unable to harness creativity? Also good to know.
I'm a little unclear how old school artists have "adapted" to tools like Midjourney. Perhaps you could clarify this point?
You're a bit fanboi with your adjectives.
Perhaps you meant to write "every corner will have access to a short film AI scriptwriter."
I wasn't aware that's what this world needed more of. Are we somehow short on scriptwriters?
Here's a novel thought: why about replacing "amateur directors" with AI as well? Then we could have spectacular films everyone could enjoy without the middleman.
So they will both "not feel the impact as much" and they will "adapt". Again, I'm unclear what you mean by "adapt".
I'm guessing it means "stop being all luddite about new technology" and "embrace AI". But then, what will AI train on?
So in the future, the AI can provide me with MIDI files so I can create my own music?
What a brave new world!
RickOli
New Member
Are you nervous? Stay calm.
It demonstrated superiority in quality and creativity, as evidenced by the number of professional composers who agree, which you can see here:
What I meant was that the designers tried to be more creative, even surpassing Midjourney and others in several aspects.
The phrase 'Screenwriters on every corner' is meant IRONICALLY.
'Adapt' means to 'BE MORE CREATIVE.' If you think AI is surpassing something, then we need to study more.
Artists with many followers will always have a loyal audience regardless of AI, so they don't need to worry.
Regarding MIDI files, there is already a plugin for that; there is no need for AI to evolve in this area.
The more people like you despair, the more people will be afraid of AI, so remain calm.
David Cuny
Grand Poobah, Royal Order of WordBuilders
This is a forum for people interested in virtual instruments. An AI trained plugin that could generate expressive performances from MIDI is literally the thing people on this forum would be interested in.
MarcusD
Tinkerer
Seems Fishy to me… maybe the next album will be using Casper AI - the not so friendly ghost writer.
Trash Panda
Clueless nitwit
My kingdom for an instrument oriented version of SynthV.
Daryl
Senior Member
So far there has been nothing that sounds particularly creative to my ears. All I hear is well worn cliches strung together, and not strung together very well either. However, even that will be good enough to put the people scrabbling about at the bottom of the industry out of work. Not necessarily a bad thing, in some ways, but it's by "scrabbling around" that people learn enough to drag themselves up to a higher level, and if this learning process is removed for us, then how is anyone going to get the experience to get to a level where they can exercise the creativity that these AI models lack?
Just a bit of thinking out loud here:
I think the future we are looking at for media creation of TV and film will be a single human "creator", or "director"...
This could still be a very creative operation, involving human writing ideas and choices and in that case the creation can still bring interesting products (we cannot go any worst then most of the garbage that came out in the last 10 years for example).
It could also be automated completely by the AI and ... I have no clue what the results may be but that is a possibility I can see easily happening.
That person will have the initial movie idea (or will scrub it from a ChatGPt like app) then will proceed to produce and direct it.
That person will prepare a script (probably still helped by AI, or not)
There will be a casting process from millions of available "actors" that can be used in the movie.
Some actors will be freeware and others will be available to download and "license" by some company that trained it's software on the best human actors (those still left or mostly past movie stars).
The "director will then proceed to describe (to prompt) every scene. Will give instructions about the "set", being able to try ideas on the fly .. ("you know what, let's make the scene happen in Paris, not Mars... oh wait try Vegas!"). And detailed instructions to the "actors" about how to carry the lines etc etc...
After the scene creation the editing process similar to what we have now but every take to choose from will be generated in real time for various options.
Special effects is the last problem obviously. "I want Gruce Billis to jump from the building onto his bike....no wait on to a T-Rex, much more fun"
And then the music. Prompt this, generate that, make that shorter, hit this and that sync point, make that part more intense, less intense, try different instruments.
The end, rolling credits: movie by Greven Stilberg
One person could do it all. If that person is very creative and interesting, has ideas and stuff to say, some artistic goals can still be met in order to create an interesting product.
Unfortunately it is very probable that the whole process will just be taken care by an AI.
The same production principal will probably be true to other media products such as videogames.
Big studio quality games created by one person while seated on the toilet ...
So this seems inevitable for me, the only irony is that whoever is starting to gleefully look to profit from having to do without graphic designers and now musicians will eventually lose their job too as their own job can be done by a single lonely prompter, or just by the software without any human input.
It's all terrifying, but genuine artists will hopefully still find a way to express themselves and touch people's feelings.
I think the future we are looking at for media creation of TV and film will be a single human "creator", or "director"...
This could still be a very creative operation, involving human writing ideas and choices and in that case the creation can still bring interesting products (we cannot go any worst then most of the garbage that came out in the last 10 years for example).
It could also be automated completely by the AI and ... I have no clue what the results may be but that is a possibility I can see easily happening.
That person will have the initial movie idea (or will scrub it from a ChatGPt like app) then will proceed to produce and direct it.
That person will prepare a script (probably still helped by AI, or not)
There will be a casting process from millions of available "actors" that can be used in the movie.
Some actors will be freeware and others will be available to download and "license" by some company that trained it's software on the best human actors (those still left or mostly past movie stars).
The "director will then proceed to describe (to prompt) every scene. Will give instructions about the "set", being able to try ideas on the fly .. ("you know what, let's make the scene happen in Paris, not Mars... oh wait try Vegas!"). And detailed instructions to the "actors" about how to carry the lines etc etc...
After the scene creation the editing process similar to what we have now but every take to choose from will be generated in real time for various options.
Special effects is the last problem obviously. "I want Gruce Billis to jump from the building onto his bike....no wait on to a T-Rex, much more fun"
And then the music. Prompt this, generate that, make that shorter, hit this and that sync point, make that part more intense, less intense, try different instruments.
The end, rolling credits: movie by Greven Stilberg
One person could do it all. If that person is very creative and interesting, has ideas and stuff to say, some artistic goals can still be met in order to create an interesting product.
Unfortunately it is very probable that the whole process will just be taken care by an AI.
The same production principal will probably be true to other media products such as videogames.
Big studio quality games created by one person while seated on the toilet ...
So this seems inevitable for me, the only irony is that whoever is starting to gleefully look to profit from having to do without graphic designers and now musicians will eventually lose their job too as their own job can be done by a single lonely prompter, or just by the software without any human input.
It's all terrifying, but genuine artists will hopefully still find a way to express themselves and touch people's feelings.
In the end, A.I. can be a formidable multimedia creation tool.
The idea of being able to create a feature film from the comfort of my home is quite seducing.
I see a difference between video generation and music generation though.
Prompt to video makes sense because it's a direct thing and not an interpretation of the prompt like it is for music.
Also, individuals can already create music on their computer (and it's really their creation) while nobody can generate a full motion video with actors alone. This is why video generation is more interesting to me (even though, again many will lose their jobs).
While this is sad for us who like the process of making music this is also very exciting for all the creative possibilities it generates.
vitocorleone123
Senior Member
Nobody bothers to listen to the music I create now! Ha.
As a hobbiest, things improve with AI tools (as long as I conveniently ignore the impact on everyone else… which I personally can’t do).
If music is a hobby and you have money, get stuff to make music. If you’re looking to anything in the arts to make a livable career out of…. Well… I hope you get lucky and either win powerball or become the next Zimmer as the odds will likely be similar.
arafaratanran
Active Member
What?
Yes, that is one of the reasons why it IS a bad thing! Sometimes I really wonder why people are not even feeling bad for the arrogance they represent in their words ... I guess this AI thing will sooner rather than later also hit those who consider themselves "upper class" and then the arrogance will hit them back like a boomerang.
All I have to say regarding AI is this: Technology started out as something to take over work from human beings that most people don't want to do. Now, we are getting to the point where technology will take over things that actually can be self-fulfilling work. Most human beings will be left to intellectually degrade and serve the interest of the few people with the funds to stay in control. As long as this continuous enslavement is not starting a revolution of some kind, this will lead to the end of progress for mankind.
pulpfiction
Active Member
I know it's a drop in the ocean...
How about we start promoting our music with a certain logo, AI-free or something like that...preferably all the same logo...
It wouldn't change much in the overall situation, but maybe a bit of promotion for our music and a bit of creating awareness on this topic...
How about we start promoting our music with a certain logo, AI-free or something like that...preferably all the same logo...
It wouldn't change much in the overall situation, but maybe a bit of promotion for our music and a bit of creating awareness on this topic...
