Advice on new income streams

[Marbury]

This is probably futile but here goes. For the past 15 years I have made a reasonable income on selling sound effects, music and images (I was a photography student 30+ years ago)

There have always been fluctuations in income but this year has absolutely tanked, this month going off a cliff.

Companies that started with good intentions for the artist to get a fair deal have since been taken over by bigger companies who do not care, and have reduced our royalty rates. I won't mention them but you see them listed at the end of most tv programmes. Add to that the fact they have all gone to subscription thus reducing what once earned Pounds to pence (in my case dollars to cents as the main royalty free libraries I am with pay in US dollars) Add to that the drop in currency from dollars to pounds and AI and it's a disaster. I have one library that owes the artists but are no longer paying or responding to messages despite the tracks still being live on their site. Again, it's in the US so no way of recouping for me.

PRS payments are also falling rapidly.

At least the mortgage is paid and my lifestyle is not decadent but bills etc are rising. If I had the contacts it would be great to work direct with clients but in this business it's who you know.

I genuinely do not know where to go from here as there are literally no options to pursue. Is anyone else in a similar boat ?

[RichardT]

Thanks goodness I'm not, but I feel for you in your current situation. I don't think it's going to get better as generative AI gets more and more powerful.

I think pretty much any sound attached to moving images is shortly going to be generated by AI, including the music - apart from the top of the market where the players are already established.

In general, my guess (no more than that) is that any area where the music or sound is secondary to something else is going to be taken over by AI.

If you can think of an area you can contribute where the sound is primary, I think there will be more chance for humans to continue to have a role.

[shufflebeat]

Folk music coupled with delivery driving.

[Arpangel]

Forget AI, the writing was on the wall ages ago, a friend who was earning great royalties from his music enabling him not have a day job so he could do nothing but make his music, but that vanished as soon as the internet and free downloads made an appearance, but hey, we can’t stop progress, can we?

[Drew Stephenson]

As I saw on threads a few months ago, "Are you even a professional musician if you don't have a course to sell on youtube?"

Unfortunately Marbury, I think you and other composers in your field are going to be right in the cross-hairs of AI generated music.

About the only thing I can suggest is to reduce the availability of your content, double your rates, highlight the hand-made nature of your work and pitch the highest tiers.
(I call this the 'Citrix' tactic after a company we worked with who provided remote desktop services. They were struggling with getting enough work and maintaining a high enough margin, so their CEO took a gamble and doubled their day rate. He made no other changes at all. Margin increased (obviously) but their order book and contracts increased as well, especially with blue-chip companies.)

[Marbury]

Unfortunately lazy people with no talent can literally use AI to generate any track to fool libraries to submit in a few seconds. I did a test with Suno earlier and asked for an experimental musique concrete electronic track from the 1950s and it was astonishingly accurate.

[Martin Walker]

Sad but true

[Arpangel]

Unfortunately lazy people with no talent can literally use AI to generate any track to fool libraries to submit in a few seconds. I did a test with Suno earlier and asked for an experimental musique concrete electronic track from the 1950s and it was astonishingly accurate.

If it sounded like that music, it was a failure.
It was a template, the exact opposite of the philosophy that made that music.

[shufflebeat]

Unfortunately lazy people with no talent can literally use AI to generate any track to fool libraries to submit in a few seconds. I did a test with Suno earlier and asked for an experimental musique concrete electronic track from the 1950s and it was astonishingly accurate.

If it sounded like that music, it was a failure.
It was a template, the exact opposite of the philosophy that made that music.

Unfortunately, what art means to those who take a critical interest is not as important as what it means to “passers-by”, there are so many more of them and they are easier to communicate with on advertisers’ terms.

[Arpangel]

Unfortunately lazy people with no talent can literally use AI to generate any track to fool libraries to submit in a few seconds. I did a test with Suno earlier and asked for an experimental musique concrete electronic track from the 1950s and it was astonishingly accurate.

If it sounded like that music, it was a failure.
It was a template, the exact opposite of the philosophy that made that music.

Unfortunately, what art means to those who take a critical interest is not as important as what it means to “passers-by”, there are so many more of them and they are easier to communicate with on advertisers’ terms.

That is very true, but I'm not sure many "passers-by" would be interested in hearing Morton Subotnic, or John Cage.

[Marbury]

It's probably too late in the day but it's a shame we can't set up some sort of cooperative site to sell direct.

[ManFromGlass]

Producers and directors don’t always have the right words to describe music so a new profession should/probably has developed.
Ai Music Wrangler.
You spend your day translating producer speak to music Ai speak.
It’s probably happening in the Ai visual world too.

[Drew Stephenson]

I believe they call themselves "prompt engineers" - which generally leads me to ask what discipline their engineering degree is in and where they graduated from.

[amanise]

It's probably too late in the day but it's a shame we can't set up some sort of cooperative site to sell direct.

Wasn't that what Bandcamp was supposed to be?

[Drew Stephenson]

It's probably too late in the day but it's a shame we can't set up some sort of cooperative site to sell direct.

Wasn't that what Bandcamp was supposed to be?

Which works as a direct-to-consumer platform but I don't think it would tick the boxes as a mechanism to pitch to filmmakers or other users of library music.

[BWC]

I believe they call themselves "prompt engineers"

Seriously though...



...no sorry, I can't take that seriously.

[RichardT]

It's probably too late in the day but it's a shame we can't set up some sort of cooperative site to sell direct.

I think it is too late. That market is going to disappear completely very soon. AI will watch the film, generate Foley, translate the dialogue into whatever languages are required and lip sync it to the characters, and generate the music and effects (perhaps based on some themes and prompts from a human, or perhaps based on nothing at all).

[Watchmaker]

When I ran up against the wall, I got an accounting degree. Had a surprisingly great career in that field for about 20 years. Made a ton of money. Now I own a staggering amount of toys which I refer to as my retirement fund. Still play music too.

[RichardT]

When I ran up against the wall, I got an accounting degree. Had a surprisingly great career in that field for about 20 years. Made a ton of money. Now I own a staggering amount of toys which I refer to as my retirement fund. Still play music too.

This is a great way to do things. Music becomes fun.

[Hugh Robjohns]

AI will watch the film, generate Foley, translate the dialogue into whatever languages are required and lip sync it to the characters, and generate the music and effects (perhaps based on some themes and prompts from a human, or perhaps based on nothing at all).

I suspect the only viewers of such a film will be AI too!

There will be a role for AI music in low budget TV dross, of course, but quality film and TV is a collaborative human art form, and people like working with people.

That's why so many film scores still use real orchestras rather than samples. People still find the money to do that because its just better.

[RichardT]

AI will watch the film, generate Foley, translate the dialogue into whatever languages are required and lip sync it to the characters, and generate the music and effects (perhaps based on some themes and prompts from a human, or perhaps based on nothing at all).

I suspect the only viewers of such a film will be AI too!

There will be a role for AI music in low budget TV dross, of course, but quality film and TV is a collaborative human art form, and people like working with people.

That's why so many film scores still use real orchestras rather than samples. People still find the money to do that because its just better.

I'm more pessimistic than you, Hugh. I agree that top of the market films will be done the old way, but my guess is that AI will impact what's done quite a long way up the food chain.

But to be fair, that's just a guess.

[ajay_m]

Interestingly if you read the lives of the great composers they often struggled financially. Beethoven, who loved coffee, had to count out beans individually. Many great artists other than musicians also lived hand to mouth.

I think the future is where smart people figure out how to use AI to do some of the more boring tasks but for example a film soundtrack is unlikely to be ai generated, a great soundtrack requires real creativity.

[Drew Stephenson]

What I haven't been able to figure out yet (mostly because I haven't looked properly) is whether generative AI is still rocketing up the acceleration curve or whether we're hitting the tipping point of the Pareto principle.
I.e. is it going to get staggeringly better over the next 2-3 years and we can just forget about any kind of professional career in audio, or are we approaching the point where it has done the 'easy' 80% of the job and further improvement is actually going to really slow down - leaving plenty of space for professional roles at the top.
I have read / heard arguments recently suggesting both.

[RichardT]

Interestingly if you read the lives of the great composers they often struggled financially. Beethoven, who loved coffee, had to count out beans individually. Many great artists other than musicians also lived hand to mouth.

I think the future is where smart people figure out how to use AI to do some of the more boring tasks but for example a film soundtrack is unlikely to be ai generated, a great soundtrack requires real creativity.

That’s a good point Ajay. Borodin was also a professor of chemistry, so he had the day job idea a long time ago.

[RichardT]

I came across this - some examples of what Google’s new Veo 3 can do - generating 720p video, voices and music.

https://arstechnica.com/ai/2025/05/ai-v ... we-doomed/

I find this flabbergasting. It’s come so far in just a year or so. Where it will be in 5 years time I can’t imagine.