Anyone interested in building an online AI music DAW for implementing on a FaaS cloud vendor?

[kivadour]

I've been thinking about possibilities and I envisioned this:

Autonomous AI agent tightly integrated with DAW

- Uses AudioGPT to analyze the generated audio: https://github.com/AIGC-Audio/AudioGPT and understand the features described in natural language ("make this bass darker and fatter).

Must've overlooked that github repo.

That's awesome.

It has great lit for "sentiment A.I."

You're getting waay ahead with connotations of everyone's own context-aware-personal-A.I.-assistant where...after the A.I. gets to know your idiosyncrasies it can employ "inflection A.I." where the way you said something will make for my earlier vision quoted in this thread of "dreaming in color".

See how a group reverse-engineered the tech used in speech recognition in Google assistant. https://www.tracklib.com/blog/digging-samples-ai


btw, this is new today: https://www.youtube.com/watch?v=map22OdIUp8&t=2s

Facebook research has made most of the code open-source!


It's not fair that visual artists get to have all the A.I. fun and not the aural artists too.

[BertKoor]

Synth -> zero or many effects -> mixer channel -> effects -> mixer bus -> mastering effects.

Do this for 30 tracks in parallel with 6 components each, on a buffer size of 128 samples at 48 kHz that's 180 serverless function calls 375 times per second. Nearly 70.000 for a second, on a track of 3:30 that's 14 million calls. It usually takes a musician some hundred attempts before even remotely satisfied. Now look up the cloud computing costs.

I think this recent link can loosen the concerns about the admitted large gobs of audio data processed that would traditionally post high costs on the online AI music DAW:

https://www.deepmind.com/blog/alphadev- ... algorithms

So you buy into the hype. I don't.

What AlphaDev did there was just do a minor optimization in an existing algorithm. The result was not x times faster. It was a minor optimization that a modern compiler could have done without you knowing about it as well.

It does not (and can not) change the structural problem of today's cloud computing costs.


The "A.I." sentiments these days reminds me of the times I grew up in: the mid seventies. Back then, when they labelled a product "digital" it then was automatically regarded as being extremely fast and precise, and attributed with supernatural capabilities. Dyson still does that. Their hair dryer has a "digital motor". Whatever that might be, they don't care to explain what that means.
In Dutch we call it "baked air".


Side track:

If you are paying peanuts, then don't be surprised you're employing monkeys.

[vstdls]

So....have you registered a taxID yet, and how much again are you willing to pay for a team to do all the heavy lifting?

[kivadour]

Synth -> zero or many effects -> mixer channel -> effects -> mixer bus -> mastering effects.

Do this for 30 tracks in parallel with 6 components each, on a buffer size of 128 samples at 48 kHz that's 180 serverless function calls 375 times per second. Nearly 70.000 for a second, on a track of 3:30 that's 14 million calls. It usually takes a musician some hundred attempts before even remotely satisfied. Now look up the cloud computing costs.

I think this recent link can loosen the concerns about the admitted large gobs of audio data processed that would traditionally post high costs on the online AI music DAW:

https://www.deepmind.com/blog/alphadev- ... algorithms

So you buy into the hype. I don't.

What AlphaDev did there was just do a minor optimization in an existing algorithm. The result was not x times faster. It was a minor optimization that a modern compiler could have done without you knowing about it as well.

It does not (and can not) change the structural problem of today's cloud computing costs.


The "A.I." sentiments these days reminds me of the times I grew up in: the mid seventies. Back then, when they labelled a product "digital" it then was automatically regarded as being extremely fast and precise, and attributed with supernatural capabilities. Dyson still does that. Their hair dryer has a "digital motor". Whatever that might be, they don't care to explain what that means.
In Dutch we call it "baked air".


Side track:

If you are paying peanuts, then don't be surprised you're employing monkeys.

It's great that you know so much about audio on a classic machine!

[BertKoor]

Anyone interested in building an online AI music DAW for implementing on a FaaS cloud vendor?

Apparently not

[kivadour]

Just stopping in to share some news to the contributors here in the thread.

DeepMind added its other new algorithms to Abseil, an open-source collection of prewritten C++ algorithms that can be used by anybody coding with C++. These cryptography algorithms compute numbers called hashes that can be used as unique IDs for any kind of data.

https://www.technologyreview.com/2023/0 ... 0of%20data.

That's for conventional servers that exist at TOP.

https://youtu.be/xbSC7ysJ1OE

Also if anyone knows how to boot trolls just PM me.

[DJ Warmonger]

can be used by anybody coding with C++. These cryptography algorithms compute numbers called hashes that can be used as unique IDs for any kind of data.

I assure you anyone coding in C++ knows what hashes are.

The news is all about the fact more efficient algorithms were discovered by AI.

[BertKoor]

20230629_184559.jpg

[prodbyjk]

AUTO-ADMIN: Non-MP3, WAV, OGG, SoundCloud, YouTube, Vimeo, Twitter and Facebook links in this post have been protected automatically. Once the member reaches 5 posts the links will function as normal.

ive been thinking of attempting something similar ala wav tool esq. im down to work on it for sure.i also wanted to create a program tha t uses an ai model to create splice astra presets based on text prompts since the xml file are pure parameter name: number mappings. feel free to email me prodbyjk@thscreative.art (mailto:prodbyjk@thscreative.art)

[Bunny_boy]

Is this thread like some sort of interactive story?