You are welcome. To understand what our principles are with regards ti AI, please have a look at aiformusic.info . We are ONLY interested in AI applications that responsibly enhance/augment human creativity. And you can hold us to this.

I will seek confirmation from our partners at Neutone about the initial one-time model training process, but as this is done with only 1-4 hours of audio source material, and the resulting models average about 30Mb, it is not going to be a resource-intensive process.

This is a fair question. Roland is committed to pursuing sustainable inferences of AI, and in this case, everything is run on-device, so using Project LYDIA requires no pervasive connectivity to a data center.

I understand the point, but I do still wonder. I am peppered with visuals I don't really care about all day long, but when I see what I believe is an AI-generated human, it stands out and I respond negatively. We will see.

Yes, I wonder, too. And short of that, if it just becomes easier to recognize, through some form of labeling, where the line between human and AI creativity might be in a given piece of music (if it exists at all- my focus was more on music created entirely with AI), I wonder how we might respond?

That's a really interesting take, thank you. I wrote a while back about the difference between emotional response (arguably to content) and emotional connection (to art, in your example). I may need to think about a follow up. Here's the first piece: twentysixtyone.substack.com/p/can-ai-cre...

It’s a good point. I think what we may yet see is blockchain and NFT become enabling infrastructure in one or more future provenance/attribution systems.

Thanks for choosing Roland Cloud! I think you will have much fun exploring the past and present, and you can look forward to the future!

And in the foreground you are seeing the read of a Roland System-100m 5 module rack.