NAMM 2026: If you’ve already made up your mind that AI music-making assistants are A Bad Thing then you might want to look away now, because LANDR has just announced two more of them.

Known as Blueprints and Layers, these are designed to assist at the ideas stage of music creation and to help you to develop those ideas into complete productions. How much heavy lifting you hand over to the AI is, of course, entirely up to you.

Both Blueprints and Layers incorporate multiple specialist AI models and were developed within what LANDR calls its “Fair Trade AI framework”. What this means, it says, is that “all models are trained on ethically sourced data from artists who opt in and are compensated for their contributions.”

As its name suggests, Blueprints is the songwriting assistant. This can generate multitrack song starters based on selectable musical parameters, such as genre, vocal style and instrumentation. As the creator, you can also provide lyrics and reference audio, and you can rework or replace any elements you’re not happy with. Once you’re satisfied, stems can be exported for use in any DAW.


There’s another option, though, which is to give your Blueprint project to Layers. Co-developed with Aiode, this can ‘listen’ to it (or, indeed, any other musical content) and augment it with any other parts you might want.

LANDR describes it as like having a roster of session musicians to call on whenever you need them, with different models covering guitar, bass, synths, horns, drums and more.

Layers can also take any sample or recording and turn it into a full stem that we’re told will work with any arrangement. “Any sample can now fit any song,” apparently.


full article: https://www.musicradar.com/music-te...ill-using-them-leave-you-feeling-empty-inside

 

Yeah, in a extreme devaluated music industry I'm sure they have obtained the dataset via the most ethical ways possible.

 

The AI boom will fail due to energy issues.

Unless there are gigantic technological leaps that dramatically reduce the energy consumption of AI chips, "the energy consumption of AI will reach a limit in the foreseeable future that is no longer socially sustainable – probably within the next ten years."

As fascinating as I am personally by the technology behind Large Language Models (LLMs) and their development, and as much as I recognize specific use cases where a trial deployment makes sense, I've been disillusioned for years by the exaggerations of the marketing and AI proponents who fueled the enormous hype. Anyone who took a sober look at the facts could see years ago that a huge bubble was building up, one that was bound to burst eventually.

But as with any bubble, people who urged caution were labeled as crackpots, naysayers, and worrywarts. I still remember the comments using the example of the steam engine or the early days of this blog. Those who didn't jump on the bandwagon would be left behind.

The exciting question for me is when and how the AI bubble will burst, what disruptions this will trigger, and what will ultimately remain, and what can be meaningfully utilized. What is already apparent, however, are the disruptions that the hype has already triggered in the job market (due to uncertainties), in companies (which are currently failing with hastily introduced AI solutions), and in the costs of storage (this will massively slow down some developments over the next two or three years).

https://borncity.com/blog/2026/01/21/sehen-wir-gerade-erste-zeichen-dass-die-ai-blase-platzt/