Hacker News Reader

Comment by dzink | Hacker News Reader

There is a very strong music copyright system in the US and in the Tom Waits case they had both music and voice impersonation. In Sky situation a person who sounds similar to Scarlet also sounds similar to some % of humanity as well. But no creative work of Scarlet was infringed upon. You can’t protect just a voice. People are born with it and if that was possible, many people on TikTok would liable for sounding like some celebrity. Being a celebrity doesn’t entitle you to a share of income from anyone who sounds like you and broadcasts.

View on HN

> There is a very strong music copyright system in the US and in the Tom Waits case they had both music and voice impersonation.

The article explicitly mentions the Tom Waits case did not include a copyright component as he didn't own the copyright, and Frito-Lay probably obtained synch rights from the copyright owner.

Tom Waits won the suit without invoking copyright violations - read into that what you may.

Bette Midler won a similar case where Ford licensed a Midler track and got a sound-a-like to perform it for a commercial.

If OpenAI's lawyers thought they were going to get away with it then OpenAI needs to get better lawyers.

> then OpenAI needs to get better lawyers.

Part of being a good lawyer is sometimes "well, this didn't work in 1992 (or whatever) but it might fly now. Let's try it, worst case is X". And, as always, the devil is in the details.

Huge disanalogies between the Midler case and this. Certainly not clear in its applicability.

Rick Astley v. Hauri is a recent case in the same vein: https://unicourt.com/case/ca-la23-richard-rick-paul-astley-v...

This one is unique because Hauri (aka "Yung Gravy") is another music artist. The Midler case is actually mentioned in the court filings, too.

It's interesting to think about the blurry line between art and commercial products.

Huh TIL, I like the song in question interestingly enough (Yung Gravy - Betty) but didn't know that the Rick Astley bit wasn't actually his voice.

I didn't either until I heard about the lawsuit!

> You can’t protect just a voice.

I'm pretty sure this is exactly why TFA was posted—apparently you can, in fact, protect a voice under California law. Of the $2.6 million in damages awarded in TFA $2 million were for "voice misappropriation".

This seems to be the relevant section of the civil code:

https://codes.findlaw.com/ca/civil-code/civ-sect-3344/

If you made me take the Pepsi challenge, I probably would not be able to distinguish most similarly-aged Hollywood actress's voices from one another in a blind test. No matter how distinct they start out, hanging around LA seems to turn their voices very similar: Young, white woman, slightly Californian accent, singsong vocal inflections, a little valley girl, a little vocal fry... After a few years, a lot of them kind of start sounding the same.

I wouldn't be surprised if there were literally thousands of voice actors that sounded close enough to any given actress to pass as her.

They didn't infringe on a Tom Waits song, they got an impersonator to sing an entirely novel song.

My take is that for a certain cohort of AI hypersters and Sam fans, it is literally impossible for OpenAI to do wrong. Sam could probably shoot a man in cold blood on national television and we'd have somebody explaining how technically it's legal because reasons.

And similarly, some people are always hungry for an opportunity to villanize OpenAI. This kind of polarized, black-and-white thinking is not productive.

Neither is false balance or enlightened centrism. The point of the GP is that we shouldn't worshipping these companies. They're not your Lord.

Yes, and the point of the "enlightened centrist" is that bad actions by a company don't mean that any complaints thrown at them are automatically valid, or that every employee at the company is acting maliciously. Pledging your allegiance to a megacorp is bad, but thinking that the court of public opinion is always correct is just as foolish.

We're not talking about any complaint, we're talking about a specific complaint. In the other thread there are about a dozen OpenAI fanboys saying the OpenAI copycat is different from the Ford copycat for a reason which is refuted by the Tom Waits case; the Tom Waits impersonator wasn't singing a Tom Waits song, they were singing about chips.

I'm not. Your comment said that, for a certain group of superfans, OpenAI can do no wrong and will always be biased in their favour. nsvd added that other people want to see OpenAI fail and will view anything regarding them negatively. I'm just elaborating on that. For one reason or another, this discourse has become personal for many people - the specific complaints don't matter, many people will be coming in with preconceived notions no matter what.

The claim is that Sam fanboys would defend Sam in any circumstance, not that Sam is guilty in any circumstance. The distinction shouldn't require explaination. If you're confused, then ask the chatbot to explain it to you.

OpenAI hasn't made it hard. Releasing ChatGPT has made a lot of peoples' lives worse, and their bait and switch from non-profit to profit are just the tip of the iceberg.

Wait whos lives are made worse by ChatGPT? I have not yet seen a claim like this

Teachers, for starters.

Ok I can see that teachers cant tell if a student has written an essay or used ChatGPT but how is that materially different from students paying for other people to write their essays? Its cheaper and easier? I knew students who payed for essays well before ChatGPT existed.

Any other examples? Not trying to defend OpenAI just interested in this idea that it has actually made anyones lives worse.

Reply on news.ycombinator.com