Hi, Everyone.

This is an article from a few days ago, but I'll post it because it's news related to music (Audio) and AI. (AI-related threads may be @tzzsmk's exclusive patents. If so, I apologize.)

“Stability AI and Arm announced on March 3 (local time) that they have developed a technology that can generate audio by running the AI model "Stable Audio Open" on smartphones. On a smartphone with an ARM CPU, it can produce audio 30 times faster.”

(https://analyticsindiamag.com/ai-ne...ring-offline-generative-audio-to-smartphones/)
(https://stability.ai/news/stability-ai-and-arm-bring-on-device-generative-audio-to-smartphones)
(https://www.arm.com/company/success-library/made-possible/stability-ai)

This could be diverted to vocals and music generation in the future.

 

Last edited: Mar 9, 2025 at 1:46 AM

impressive, but also wasted potential, because vast majority of audio folks don't run their DAWs on phones

 

I see.
Thank you for your helpful input.

There are many people writing songs on iPhones and smartphones today, as exemplified by Steve Lacy.
I also believe that many of the younger generation are more accustomed to handling smartphones than PCs.
Moreover, by connecting a mobile device to a PC, it is possible to transfer the audio samples and other data inside the smartphone.
In case you are wondering, Stable Audio Open is a service that can also be used on PCs.
I have decided that although the service is still in its infancy, it would not hurt to keep an eye on future developments.

 

Please remember that you are increasingly handing over your self-determined actions to a machine. In addition, the AI software will severely limit your creativity. True to the motto: Drawing a picture yourself is exhausting and takes time, the AI generates a picture for me quickly and in a matter of seconds.

People are lazy and curious by nature and therefore the new tools will also be successful with people who are uncritical of AI programs.

Have you ever had the idea of really thinking and reading about what technology does to you and your brain or are you so naive that you believe that everything that is available to buy and use is also good for people.

The big tech companies will only tell you the advantages of their product but not the disadvantages, because they want to sell their products in large quantities.

 

indeed I've tested Stable Audio Open locally a lot - it's fun especially for sound design (like busy streets, vehicle crashes, water dripping in caves etc...) or sample generation (meaty kicks, shitty snares etc..),

my point is, audio - just as video - is very difficult to describe by mere text - I've started testing local video generation and as a matter of fact results are much better if input is a handpicked (generated) image,
I'm thinking of audio the same - I've recently tested lyrics-to-vocals models accompanied by (AI-split) instrumental backing tracks, quality is already much better that way compared to going all in from scratch