Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Google's advanced music generation model and two new AI experiments

deepmind.google

Post reply on HN
Print view View original

171–180 of 361 posts

Re: Google's advanced music generation model and two new AI experiments

#171

Post by Musikstueck »

InCityDreams wrote: post #123
Musikstueck wrote: post #108

What we do as humans is to push against a wall of mist of a unlimited search space. Doesn't matter if it is music, drawing or scientific discoveries. What an Artist does is pushing the boundaries. Having tools like ths allows much more people to push and not 'just' the artists who mastered certain skills. I can write very good software, do i care if my family suddenly can also write softwre thanks to ai? Honestly no.…

>Doesn't matter if it is music, drawing or scientific discoveries. #oxford comma blues.

Im german, had the second worst grade in english when i graduated. Feel free to be more precise than a hashtag :P

Re: Google's advanced music generation model and two new AI experiments

#172

Post by kmfrk »

> SynthID embeds a watermark into AI-generated audio content that’s inaudible to the human ear and doesn’t compromise the listening experience. It does this by converting the audio wave into a two-dimensional visualization that shows how the spectrum of frequencies in a sound evolves over time. This novel method is unlike anything that exists today, especially in the context of audio.

Is this a unique or boolean watermark (==ai_generated)? Sounds like it would be amazing to have for regular YouTube uploads to prove authorship and deal with reuploads/content theft.

Re: Google's advanced music generation model and two new AI experiments

#173

Post by fodkodrasz »

seydor wrote: post #82

Wonder where the human fits in this new landscape. Composer? no. Performer? no. Music artist is more like a conductor now. Thankfully audience doesn't care because they need to worship a human in the end, so she 'll be receiving all the money anyway. The problem with these new BigTech companies is that they are accustomed to subjugating everyone. They deal with users and content, not with artists and creations. This…

> Thankfully audience doesn't care because they need to worship a human in the end

Not the HN audience. Luckily in the background there will be a machine or a even better, a company that can be worshipped!

Re: Google's advanced music generation model and two new AI experiments

#174

Post by kevindamm »

tchaffee wrote: post #143
seydor wrote: post #85

Earlier quoted context omitted.

Hm. It was before the 20th century maybe. People knew of artists by name, not by their social life and sex tapes. The pop culture paraphilia is a very recent thing.

You're claiming that gossip and hero worship and celebrity didn't exist before the 20th century which is a wild claim that you'd need to back up with evidence. Humans haven't changed that much over the centuries.

..but our ability to communicate quickly and broadly has.

Re: Google's advanced music generation model and two new AI experiments

#175

Post by phatskat »

darkerside wrote: post #156
tchaffee wrote: post #128

Earlier quoted context omitted.

Now you need to define what a unique, profound emotion is. And you need to define great execution. And you need to explain why intentionally having poor execution couldn't express a unique, profound emotion. On Radiohead's biggest hit Creep "That's the sound of Jonny trying to fuck the song up. He really didn't like it the first time we played it, so he tried spoiling it. And it made the song." At the end of the day…

That true? Haven't heard that story

Quick google and found this[0] - fun little read so far

[0] https://www.theringer.com/2021/11/17/22785868/radiohead-cree...

Re: Google's advanced music generation model and two new AI experiments

#176

Post by microtonal »

datpiff wrote: post #144
jacquesm wrote: post #100

Earlier quoted context omitted.

Arguing a general case by pointing out a massive exception doesn't really work though.

The guitar is literally the most mass-produced instrument in human history

I was quite surprised by this. I played some guitar in my teens and a good guitar was quite an investment (mid-late nineties). My dad is a bit of a guitar collector and it is surprising what guitars he finds for ~300 Euro, quite respectable quality.

(Maybe it also has to do something with the lack of demand now, guitar music was still very popular in the 90ies.)

Re: Google's advanced music generation model and two new AI experiments

#177

Post by sesm »

tchaffee wrote: post #150
sesm wrote: post #127

Earlier quoted context omitted.

I have hard time believing that you can learn quicker playing chess against AI. Humans can explain ideas behind their moves, their long term plans and their position evaluation. That what chess coaches will explain you and that's what you can read in chess books. Computer has hard time explaining his moves, has no notion of long-term plan and it's position evaluation is often useless for beginner or intermediate ches…

Some of what you are claiming is false. Most chess AIs for example can identify almost every opening move in existence. Something no human can do. That's a type of position evaluation, and a good starting point for further research. No free coach is going to sit for 4 hours straight, 5 days a week, playing thousands of games exploring variations of a particular opening. That kind of self-study will definitely help yo…

Chess engines are very important for grandmasters, for sure. But for beginners they are mostly useless. That’s what I was arguing about.

Regarding openings - the best place to learn are books/studies/even Wikipedia articles that explain the ideas behind the major moves and variations, which chess engines don’t do. Also, a human can explain which opening are good to play at your skill level and which are not.

Regarding free coaches, playing with a friend/relative over the board and discussing the game afterwards has a similar effect.

Re: Google's advanced music generation model and two new AI experiments

#178

Post by pcdoodle »

I never want to hear this stuff but you know they're going to make it available for every content junkie publishing hype beast low effort mouth open consumer product placement having youtuber so it's going to be forced into your ear holes at some point. I hate google so much.

Re: Google's advanced music generation model and two new AI experiments

#180

Post by voidhorse »

bambax wrote: post #64
halflings wrote: post #41

Earlier quoted context omitted.

Growing pains. Remember how this all started (Google Magenta, circa 2019): https://www.youtube.com/watch?v=ZRnbbtqxBEc

It seems most (or all?) AI music projects aim at producing audio? Why not output MIDI instead, and let the artist manipulate that? I for one would gladly buy a VST instrument that would generate a constant stream of MIDI ideas, chordings, voicings, variations, from a few seconds of singing or whistling or clapping. It seems the technology is here to build it, yet (AFAIK) it doesn't exist. Why?

Its because the ultimate goal with these gen AI tools is to replace not support professionals. Your idea makes sense, is eminently doable, and would be awesome. The problem is, it still requires a human with expertise. Gen AI companies want to remove the human with expertise from the equation precisely so they can "empower" (take money from) the much wider population of humans without expertise. These humans, since they lack expertise will actually need gen AI to produce anything at all. A professional could use it but could ultimately make do without it--they would not have a strong dependency on gen AI. It's essentially the same idea as addicts being lucrative prospects for drug dealers--once you establish dependency, you've got them.

Let's be frank, gen AI projects are ultimately about companies wanting make money by putting small-scale professionals out of work for good. In other words, they want to capture a significant portion of all that capital that currently flows across small companies and projects (even if small scale operations continue a large percentage of their production will require funneling money to Gen AI subscriptions instead of skilled human workers)

Post reply on HN