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

Stable Audio Open

stability.ai

Post reply on HN
Print view View original

121–130 of 137 posts

Re: Stable Audio Open

#121

Post by jrm4 »

Workaccount2 wrote: post #87
jrm4 wrote: post #57

Earlier quoted context omitted.

I fail to see how the argument is ridiculous; and I'll bet that a jury would find the idea that "there is a copy inside" at least reasonable, especially if you start with the premise that "the machine is not a human being." What you're left with is a machine that produces "things that strongly resemble the original, that would not have been produced, had you not fed the original into the machine." The fact that there…

If a made a bot that read amazon reviews and then output a meta-review for me, would that be a violation of amazon's copyright? (I'm sure somewhere in the amazon ToS they claim all ownership rights of reviews). If it output those reviews verbatim, sure I can see the issue, the model is over fitting. But if I tweak the model or filter the output to avoid verbatim excerpts, does an amazon lawyer have a solid footing fo…

Generally yes. You're talking about "derivative works."

Re: Stable Audio Open

#122

Post by GaggiX »

jrm4 wrote: post #120
GaggiX wrote: post #65

Earlier quoted context omitted.

Having exact copies of the samples inside the model weights would be an extremely inefficient use of space, and also it would not generalize, unless it generated a copy so close to the original that it would violate copyright law if used, I wouldn't find it very reasonable to think that there is a memorized copy inside the model weights somewhere.

But you're a tech person. I'm trying to think of this from the point of view of e.g. a likely potential jury. Again: Imagine two AI machines, different in one way: One of them has been fed "Article X" and the other hasn't. You press buttons on the machine(s) in the same way. The machine that was fed "Article X" spits out something that looks like Article X, and the one that wasn't, doesn't. The magic inside, I don't…

>But you're a tech person. I'm trying to think of this from the point of view of e.g. a likely potential jury.

Courts can call experts to testify on matters requiring specialized knowledge or expertise.

Re: Stable Audio Open

#123

Post by hecanjog »

blargey wrote: post #81
hecanjog wrote: post #66

Earlier quoted context omitted.

I also highly doubt anyone who signed agreements to have their music included in the Free Music Archive would have been OK with this. The particular type of license was important to contributors and there's a difference between allowing for rebroadcast without paying royalties and allowing for derivative works... I don't really care to argue the point, but it's why there were so many different types of licenses for t…

If you look at the repo where the model is actually hosted they specify > All audio files are licensed under CC0, CC BY, or CC Sampling+. These explicitly permit derivative works and commercial use. > Attribution for all audio recordings used to train Stable Audio Open 1.0 can be found in this repository. So it’s not being glossed over, and licenses are being abided by in good faith imo. I wish they’d just added a se…

That's an important clarification, thanks! The full FMA database has lots of music that isn't licensed that way.

Re: Stable Audio Open

#124

Post by domothy »

rfoo wrote: post #114
hanniabu wrote: post #82

Earlier quoted context omitted.

> Also, only ~30% tokens are staked. The 30% who chose to stake essentially tax the other 70% in use. And in PoW miners tax 100% of holders. > what they receive is proportioned to how much they stake Wealthy miners with state of the art ASICS benefit more than some kid mining at home with an old GPU. Maintenance/cost of mining equipment benefits from economies of scale too. I hate being mean, but sorry, remembering t…

Yeah, PoW is bad too. But I'm happy to pay a tax to those who burnt energy to keep the networking running and converted USD to the native token, proportioned to their effort. I'm less happy to pay someone a tax just because they are rich and they did barely anything.

>But I'm happy to pay a tax to those who burnt energy to keep the networking running and converted USD to the native token, proportioned to their effort.

Why? That's doubly bad for non-mining holders: not only does your share of supply get diluted with newly printed coins, but it also get devalued relative to USD when these coins inevitably get sold to pay expenses

In Ethereum's post-merge world, non-staking holders can keep their share of supply the same (or even have it passively increase) when total supply shrinks. And if the supply does increase by ~0.5-1% and you as a holder aren't okay with that amount of dilution, the barrier of entry to stake profitably and protect your share of supply is much, much lower than the barrier of entry of profitable bitcoin mining.

And the total newly issued coins (which are nominally much lower than pre-merge) have a much lower need to be sold off. If you view issuance as a tax on holders, Ethereum's model wins on all counts

>just because they are rich and they did barely anything

But stakers also "keep the networking running", just like miners under PoW. In both cases, it's gonna be the amount of capital involved that decides how the rewards are proportioned, there's no way around it - these permissionless systems ultimately use the inherent scarcity of economic capital as the anti-sybil mechanism with economic incentives to keep everyone honest. PoS just bypasses the need for burning a huge amount of energy and the embarrassing quantity of single-purpose e-waste to indirectly calculate who has how much at stake. It goes straight to the point: the capital at stake is simply measured in the value of the coin itself instead of external energy/hardware.

On the outside it does kinda look like stakers get rewarded passivly for doing nothing, but there are definitely costs involved, they're just mostly economic instead of physical - think of all the usual risk involved in crypto's volatility, now compound that with slashing risks, illiquidity, opportunity costs – staking yield is like 4-5% atm (and has been down only for quite some time), if you're a billionaire whale you definitely have other investment opportunities available that yield way more than that. I mean just the fact that the net supply growth can go negative shows that even internally in the blockchain itself there can be better things to do with your ETH than stake it; these people aren't burning their ETH on transactions fees for fun, they're actively using their ETH to do stuff that gets them some economic utility.

Re: Stable Audio Open

#125

Post by waffletower »

jrm4 wrote: post #57
Workaccount2 wrote: post #24

Earlier quoted context omitted.

The idea that AI trained on artist created content is theft is kind of ridiculous anyway. Transformers aren't large archives of data with needles and thread to sew together pieces. The whole argument is meant to stifle an existential threat, not to halt some illegal transgression. If they cared about the latter a simple copyright filter on the output of the models would be all that's needed.

I fail to see how the argument is ridiculous; and I'll bet that a jury would find the idea that "there is a copy inside" at least reasonable, especially if you start with the premise that "the machine is not a human being." What you're left with is a machine that produces "things that strongly resemble the original, that would not have been produced, had you not fed the original into the machine." The fact that there…

But you already have the same situation with people -- 'what you're left with is an artist that produces things that strongly resemble the original, that would not have been produced, had the artist not studied the original work'. Yes it is ridiculous.

Re: Stable Audio Open

#126

Post by rfoo »

domothy wrote: post #124
rfoo wrote: post #114

Earlier quoted context omitted.

Yeah, PoW is bad too. But I'm happy to pay a tax to those who burnt energy to keep the networking running and converted USD to the native token, proportioned to their effort. I'm less happy to pay someone a tax just because they are rich and they did barely anything.

>But I'm happy to pay a tax to those who burnt energy to keep the networking running and converted USD to the native token, proportioned to their effort. Why? That's doubly bad for non-mining holders: not only does your share of supply get diluted with newly printed coins, but it also get devalued relative to USD when these coins inevitably get sold to pay expenses In Ethereum's post-merge world, non-staking holders…

> but it also get devalued relative to USD when these coins inevitably get sold to pay expenses

Good point. It invalidates the "good" part but does not make it doubly bad IMO.

And for ETH, well, I don't think it's about protecting value, it's more about:

> In both cases, it's gonna be the amount of capital involved that decides how the rewards are proportioned, there's no way around it

Yes. The difference is, PoW requires you to BURN resource proportioned to your rewards, while PoS just requires you to HAVE (but not burn) it. This makes a huge difference IMO.

For example, I would consider it more "ethical" (whatever that means) to add a light PoW part (with constant or slowly increasing difficulty, that is chosen to reduce environmental impact) to the ETH PoS protocol as-is: the random-chosen validators have to solve a PoW in addition to make their efforts proportioned to how much they stake, instead of being mostly constant.

Re: Stable Audio Open

#127

Post by waffletower »

ziofill wrote: post #75
GaggiX wrote: post #65

Earlier quoted context omitted.

Having exact copies of the samples inside the model weights would be an extremely inefficient use of space, and also it would not generalize, unless it generated a copy so close to the original that it would violate copyright law if used, I wouldn't find it very reasonable to think that there is a memorized copy inside the model weights somewhere.

A program that can produce copies is the same as a copy. How that copy comes into being (whether out of an algorithm or read from a support) is related, but not relevant.

A human illustrator can also copy existing works. As a result they are not criminalized for making other non-copies. The output of an AI needs to be considered independently of its input. Further, the folly of copyright ought to be also considered, since no work -- whether solely human in origin (such as speech, unaccompanied song, dance etc.) or built with technological prosthesis/instrumentality -- is ever made in a cultural vacuum. All art is interdependent. Copyright exists to allow art (in the general sense) to be compensated for. But copyright has grown into a voracious entitlement for deeply monied interests, and has long intruded upon the commons and fair use.

Re: Stable Audio Open

#128

Post by hanniabu »

rfoo wrote: post #126
domothy wrote: post #124

Earlier quoted context omitted.

>But I'm happy to pay a tax to those who burnt energy to keep the networking running and converted USD to the native token, proportioned to their effort. Why? That's doubly bad for non-mining holders: not only does your share of supply get diluted with newly printed coins, but it also get devalued relative to USD when these coins inevitably get sold to pay expenses In Ethereum's post-merge world, non-staking holders…

> but it also get devalued relative to USD when these coins inevitably get sold to pay expenses Good point. It invalidates the "good" part but does not make it doubly bad IMO. And for ETH, well, I don't think it's about protecting value, it's more about: > In both cases, it's gonna be the amount of capital involved that decides how the rewards are proportioned, there's no way around it Yes. The difference is, PoW req…

Okay so startup capital aside, you like PoW because there's an ongoing cost to participating in consensus and don't like how PoS is more or less free.

I'm not sure how you arrived at this when your initial complaint was the rich get richer with PoS. PoW has much higher costs to participate and after a few years you have more costs when you need to upgrade your mining rigs because they're either burnt out or outcompeted by newer hardware.

Re: Stable Audio Open

#129

Post by jrm4 »

waffletower wrote: post #125
jrm4 wrote: post #57

Earlier quoted context omitted.

I fail to see how the argument is ridiculous; and I'll bet that a jury would find the idea that "there is a copy inside" at least reasonable, especially if you start with the premise that "the machine is not a human being." What you're left with is a machine that produces "things that strongly resemble the original, that would not have been produced, had you not fed the original into the machine." The fact that there…

But you already have the same situation with people -- 'what you're left with is an artist that produces things that strongly resemble the original, that would not have been produced, had the artist not studied the original work'. Yes it is ridiculous.

Right. But broadly, the law also strongly tends to distinguish humans from machines in this space. (Which, imho, is a very good idea)

Re: Stable Audio Open

#130

Post by Tricky_Troll »

rfoo wrote: post #114
hanniabu wrote: post #82

Earlier quoted context omitted.

> Also, only ~30% tokens are staked. The 30% who chose to stake essentially tax the other 70% in use. And in PoW miners tax 100% of holders. > what they receive is proportioned to how much they stake Wealthy miners with state of the art ASICS benefit more than some kid mining at home with an old GPU. Maintenance/cost of mining equipment benefits from economies of scale too. I hate being mean, but sorry, remembering t…

Yeah, PoW is bad too. But I'm happy to pay a tax to those who burnt energy to keep the networking running and converted USD to the native token, proportioned to their effort. I'm less happy to pay someone a tax just because they are rich and they did barely anything.

>I'm less happy to pay someone a tax just because they are rich and they did barely anything.

As the operator of a single validator node you can get out of here with that take. I'm using up very significant bandwidth, having to keep a computer running 24/7, updating node and OS software, troubleshoot it after a power or internet outage and at some point I will have to replace the SSD since it is constantly reading and writing and will need replacing after a few years.

Is it a full time job? Absolutely not, but is it free from responsibility? Definitely not. If anything, I could be making more than 3%pa elsewhere if I weren't also in it for ideological reasons.

Post reply on HN