Live data from Hacker News

AI generated music to improve focus, relaxation and sleep

brain.fm

51–60 of 307 posts

Re: AI generated music to improve focus, relaxation and sleep

#53

Earlier quoted context omitted.

Dr. Steven Novella on brainwave entrainment, from 2008: http://theness.com/neurologicablog/index.php/brainwave-entra... "That is generally where the science ends and the pseudoscience begins. A number of companies and individuals have then extrapolated from the phenomenon of entrainment to claim that altering the brain waves changes the actual functioning of the brain. There is no theoretical or empirical basis for t…

Yeah, I'd also like to see a study where they test this magic-AI-music against some generic 10 hours relax/focus music from Youtube (or the Machinarium or Age of Empires 2 soundtrack). Properly randomized and blinded etc. I'd also like to know what it means that this is produced by an AI. What do they optimize for? I mean it would be really cool if they could do some reinforcement learning with real time EEG feedback…

It's called "commitment and consistency" among Cialdini's six principles of influence (in "The Psychology of Persuasion", which I believe is sort of a classic)

Re: AI generated music to improve focus, relaxation and sleep

#55

I just worked on something similar at a hackathon this weekend with the Muse Headband, it was really fun. We used a support vector machine to maximize the "calmness" output of the band, which is some weighted sum of gamma and beta waves I believe.

What you've worked on sounds much more scientific than OP.

I'm not sure if that's fair.

They (claim to) have [0] a bunch of PhDs, and have a page with the same kind of words on [1] as parent.

  [0] https://www.brain.fm/about  
  [1] https://www.brain.fm/#science

Re: AI generated music to improve focus, relaxation and sleep

#56
post #51

As someone who experiences asmr I was expecting a somewhat similar experience while listening to the 'sleep' module/mode. But I don't think this does anything for me, in fact the sound set me a bit on edge for no apparent reason.

Not me, I'm listening to the Calm -> Guide Meditation one and wow is it relaxing. Not quite getting asmr symptoms, but one average I feel quite relaxed and content. I agree on the on edge quality, for me it's subtle though and probably why I'm not getting full blown asmr bliss.

Re: AI generated music to improve focus, relaxation and sleep

#57
I've been using music for years to improve my focus, relaxation, sleep, happyness, sadness, social relations and many more. Music make us feel things, we all know that. For me it's out of question that it can make you better, speaking broadly.

Now, having brain.fm picking the music so I can feel X seems like a step backwards. I already can do that but much better just using the wonderful universe of existing recorded music.

I don't see the value for me here.

Re: AI generated music to improve focus, relaxation and sleep

#58
post #2

Curious about the tech/science behind this. Is it mostly binaural beats?

Hi there, we don't use binaural beats. Or isochronic tones. Our CTO (Adam Hewett) is the founder of Transparent Corp. His older software is used to create the vast majority of isochronic tones, and all those videos you see on youtube, etc. Any rhythmic audio stimulus will entrain the brain. Actually, the very first research on audio brainwave stimulation used simple clicks. As another example if you had a drummer tha…

The sounds are unusually good for algo music, which typically has very poor production values.

I doubt you've perfected HRTF, because everyone has a different HRTF, so you'll always get an approximation at best.

>Imagine a rain sound for a second. Hundreds of thousands of drops! Imagine now that each and every drop is synchronized to the purpose of the stimulation. Now imagine a symphony... That is Brain.fm.

This is marketing. As a statement about technology or user benefits, it's meaningless.

To be fair, I liked the audio a lot, and it's the best algo-comp effort I've heard yet.

The literature on entrainment seems unclear. There are a few studies about binaural beats I know of. There's not a lot in mainstream neurological journals - just in a few fringey publications - although this may have changed since I last looked.

I think if you're claiming AI-created symphonies of brain stimulation you're going to have an easy sell to the New Age demo, and a very tough sell to a lot of other demos.

Post reply on HN