AI is overhyped

 

Actually, being replaced by AI in film or in the advertising field is underhyped.

 

As @The Aural Aesthete mentioned, TV series composing is much more demanding than film. In film you get more creative in a way. In TV you gotta meet the deadlines and they are always last minute things going on (at least in my area). I'm not a composer, i'm a re-recording mixer and i work closely with composers. One thing i can tell, is that we have no holidays, Sundays, etc. We talk pretty much after hours and i get new tracks at 4-5am to use in the next day's delivery. The most important part (other than my main composer is a machine , writing tracks on the spot) is the sound of it. I work with 3-4 composers simultaneously and the one who nails it everytime, is the one who has the best sound. Not the best chord progression etc. He might send a piece with just 3 chords but it sounds like a full polished production sound , even from his unmastered demo. You could even use that part and no one could figure it. It makes the production sound valuable. Anyone of us can listen for example a Steinwey-like piano with lush sound and distinguish it from a MIDI sample (even if both sounds are sampled). The end result is what counts. That is his greatest distinction from other composers and that is the same reason that Hans Zimmer is dominating blockbusters. He doesn't have any complex compositions that only a genius would compose. He has THAT sound , that sells. This is where the money is. Make your production value look like a million dollar budget polished mix and you'll get new doors open for you. otherwise you will not make a progression.

Oh and by the way, before working on a series that turned out to be good and opened new doors for him, he was almost bankrupt. This ist the part that you just need to be lucky enough to work on such a project. Then comes the PR and new doors

 

The same oxymoron has been around for decades. Which is? - You need a great CV of compositions and jobs to always get the really good work, but if you haven't got a good reputation doing this, how can you get the really good work? I agree with everyone who said it is much harder this century.

The Internet, quality sampled Orchestras and instruments at everyone's fingertips that are gradually replacing the need to hire many musicians as a couple of examples, means nearly any musician can theoretically write a decent composition and put it out to the whole world.
One thing that has not changed is when you meet the director, or producer, or author, or corporate owner...etc... Your attitude. Because the Internet provides them with an abundance of decent music and writers to choose from, how you come across will sell you, because it is likely, that unless you have discovered a brand new style that is uniquely yours that everyone wants, there are 100s of thousands who can do the job as well . Talent is cheap.

Think of it like your own music you write for yourself. It's your baby. Their film, doco or game or project is their baby they are trusting you with to write perfect music to match their vision. They may like the music but if they do not feel you are a good fit as a person, it's all over.

 

Last edited: Oct 4, 2024

It's overhyped. I don't know of anyone who has actually been replaced by AI in "real jobs" and where film/games are concerned: no one outside of the indie realm, and even there it isn't entirely replaced. It will vary on project, but what seems to be more common right now is generative AI for greenscreen backgrounds. Some people might use AI for chroma key but not music; others, music but not chroma keys.

The arts have historically been the only field in which automation increases the value of "manual labor". In the '90s, autotune was heralded as the end of real singers, but people are now only more impressed by singers who can actually sing on pitch, and the existence of autotune has not prevented them from having a career. Samples have not put orchestras out of business, because people using them were never in a position to hire real orchestras, or at least not the complete ensemble, anyway.

Indie game devs and such who couldn't afford to or didn't want to hire real composers either blatantly stole music, got some desperate n00b to do it for free, or bought premade loop packs and threw it together. I believe that was the case for "Undertale".

AI is simply another option instead of using loops or stealing stuff.

Where the big dogs are concerned: Some of the absolute biggest movies and most acclaimed ones of the last 10 years were so because of their heavy reliance on practical effects. Most recently: Alien Romulus.

As I said: Automation in the arts increases the value of "doing it for real". The same cannot be said for blue collar work. Now that robots have largely taken over milking cows, the value of doing it by hand has only declined. They make guitars in factories, but people still shell out thousands for hand-made ones.

"Theoretically", but practically it is a different story.

 

Most of what passes for music now is so terrible AI might have written it anyway, so who cares if it was written by AI or a human? It's all crap. AI just contributes to the huge amount of background noise and static that anything that had any slight glimmer of decency or humanity gets swallowed up in. Let AI take over, what honest difference is it going to make? How is 90% of what Hans Zimmer puts out any slight improvement over AI beyond the fact that he has access to a few interesting real instruments that AI doesn't?

 

Yes the entire American Actors unions went on strike to protest the possibility of it ever occurring which was Global news and why so many of people's most loved shows went on a hiatus. So there is total validity in that thought process. I am yet to meet anyone who does not have millions of dollars who would like to be replaced.

 

I saw a video of a cat standing on a keyboard set to a pad patch and it came out sounding more musical than "Dune"

 

...yet.

 

no, I cant.
I have to work on a "regular" job.

Like this

 

I dont want to make a living.

 

In the first place, the idea of making a living as a musician is wrong.

A long time ago, a certain music group suddenly announced their disbandment at the height of their popularity. When they explained the reason for their disbandment, they said the following.

"What is success as a musician? I don't really know. When I started thinking about it, I gradually became more and more depressed. Does making a lot of money from music mean that you're not doing the music you like?"

If you become famous and sell well, does that make you a successful musician? From a business perspective, the answer to this question is yes, but if you ignore the business aspect and get to the essence of what success is for a musician, the answer is no.

There are plenty of great musicians who don't have hits, and plenty of musicians who can't make a living from music but who write great songs.

My extreme opinion is that all musicians should live a life like Van Gogh.

 

Last edited: Oct 12, 2024

Anyone who used to make a reasonable side-income from stock music sites will strongly disagree with you.

Not to mention composers who were previously earning 6-figure incomes from the same sites. For me it was just a dumping ground for client-rejected tracks, but some of the music Ive heard created via AI, though not my genre or taste, has been extraordinary.

 

Without highly placed connections in the scene today, I think it is a definite no now. At my age I really don't care, I compose for my own pleasure and will never stop, I play and share my music with friends and their appreciation of my work is what pleases me the most. I don't have to compete to or impress anyone. My advice is just play with passion and enjoy.

 

sure

twice in a year

 

Ridiculous

 

There are still people making reasonable incomes from stock music sites. Maybe not "pond 5" and the like, but professional libraries with connections, yes.

 

Hmm. Perhaps youre intentionally missing the point? The fact is that composer incomes across the board through stock music sites has seen a rapid decline over the past 18 months, mainly due to clients moving into AI generated music sites. Perhaps you might point me towards one of the "professional library" sites and we can share the experiences of the composers within their respective forums?

Its not really disputable.

 

This is true in the sense that the noobs are scrambling for it thinking it's going to create wealth, make them into super producers/composers/artists with little effort and in a short span of time. This overhyping is effective for the platform owners who will sell this snake oil until they actually get it to that advertised performance. YouTubers are earning a living hyping this as "educational condent", which Google pays significantly higher ad revenue for in the US. Music/Entertainment earnings are ridiculously low on YT. Thats why you always see vids like "If you're not making $10,000 a month..." or "What's the best AI tools of 2024".

The industry giant have always sought to drive the cost of services and products they PAY for. And creative works happens to be in that target. Ai is a tool they will harness to do so. Many companies have already expressed this intent. Google also admitting that recent layoffs are related to new Ai initiatives. Just look at the free version of Copilot, it writes decent code. Imagine the premium services.

This is also true, because I have witnessed writers in the industry already using it for outlines and then building out the ideas in Final Draft software. It is happening and it will evolve.

The industry giants have always sought to drive down the cost of services and products they PAY for. And creative works happens to be in that target. Ai is a tool they will harness to do so. Many companies have already expressed this intent. Google also admitting that recent layoffs are related to new Ai initiatives. Just look at the free version of Copilot, it writes decent code. Imagine the premium services.

The recent interview with Eric Schmidt reaffirms what is coming in the future. Don't take it lightly. Enterprise Ai is a different beast from Consumer AI. Most of the innovations we have been playing with (Leonardo, DALL-e, Stable Diffusion, etc) are already just a subset small scope of what is being developed. Its like a cold war where everyone with a computer has a shot at developing something. Nvidia CEO also mentioned that in less than 20 years, we'll able to tell AI "Make me an application like ...". This will affect software development industry, but software engineers will still be around, just less of them working for a company, and more of them being entrepreneurs.


I respect everyone's opinion here, especially those with experience in the field. I like meditating on insight.

 

But seriously I'm not seeing any evidence that composer income from production libraries is dropping and specifically due to AI. It's essentially an unfalsifiable claim.

What television companies or film who were previously licensing human-made pieces are now using AI instead?

What's starting to happen is that AI is going to be scapegoated for poor sales and often of poor music. Long before AI, the majority of library songs never got placed, so it wasn't AI's fault.

Now what will happen is some guy will upload his crappy "beats", they don't sell, and he will blame AI without any evidence that he was passed up by Discovery Channel for Suno.