Live data from Hacker News

OpenAI didn’t copy Scarlett Johansson’s voice for ChatGPT, records show

washingtonpost.com

31–40 of 1001 posts

Re: OpenAI didn’t copy Scarlett Johansson’s voice for ChatGPT, records show

#31

"She worked closely with a film director hired by OpenAI to help develop the technology’s personality." So you are trying to tell me there was zero chance that this film director was not aware of the movie "Her" and may have been influenced by it? Why doesn't the voice sound like the Enterprise's computer from TNG? I don't mean sound, I mean cadence, more professional and not like a sexline operator.

What's your point here? Did they or did they not use another voice actor who gave them permission to use their voice? The voice is similar to Johansson's, sure, but it's not her voice and it wasn't an AI generated version of her voice. I'm failing to see what OpenAI did wrong here.

Re: OpenAI didn’t copy Scarlett Johansson’s voice for ChatGPT, records show

#32
post #18

The thing that worried me initially was that: - the original report by Scarlett said she was approached months ago, and then two days prior to launch of GPT-4o she was approached again Because of the above, my immediate assumption was that OpenAI definitely did her dirty. But this report from WaPo debunks at least some of it, because the records they have seen show that the voice actor was contacted months in advance…

Yes, but it changes the narrative from “they couldn’t get Scarlett to record the voice, so they copied her voice” to something much less malicious. Contacting Scarlett, when you already have voice recordings ready but would prefer someone famous, isn’t that bad of a thing imho.

Re: OpenAI didn’t copy Scarlett Johansson’s voice for ChatGPT, records show

#33
> an actress was hired to create the Sky voice months before Altman contacted Johansson

> the actress confirmed that neither Johansson nor the movie “Her” were ever mentioned by OpenAI. The actress’s natural voice sounds identical to the AI-generated Sky voice, based on brief recordings

Given this I don't think anyone at OpenAI did anything wrong in this instance except Sam Altman. After getting explicitly rejected by Johansson he should not have asked again and definitely should not have referenced her character in that tweet. And this whole thing could have been avoided if they just used a different voice for the demo instead of their voice that happened to sound the most like her.

They should have learned from Weird Al. Famously, he technically doesn't need permission to do song parodies but he asks anyway and respects the artist's wishes if they say no.

Re: OpenAI didn’t copy Scarlett Johansson’s voice for ChatGPT, records show

#34
post #18

The thing that worried me initially was that: - the original report by Scarlett said she was approached months ago, and then two days prior to launch of GPT-4o she was approached again Because of the above, my immediate assumption was that OpenAI definitely did her dirty. But this report from WaPo debunks at least some of it, because the records they have seen show that the voice actor was contacted months in advance…

I don't know, to me, it's just sounds like they know how to cover all their bases.

To me, it sounds like they had the idea to make their AI sound like "her". For the initial version, they had a voice actor that sounds like the movie, as a proof of concept.

They still liked it, so it was time to contact the real star. In the end, it's not just the voice, it would have been the brand, just imagine the buzz they would have got if Scarlett J was the official voice of the company. She said no, and they were like, "too bad, we already decided how she will sound like, the only difference is whether it will be labelled as SJ or not".

In the end, someone probably felt like it's a bit too dodgy as it resemblance was uncanny, they gave it another go, probably ready to offer more money, she still refused, but in the end, it didn't change a thing.

Re: OpenAI didn’t copy Scarlett Johansson’s voice for ChatGPT, records show

#36

I was perusing some Simpsons clips this afternoon and came across a story to the effect of "So and so didn't want to play himself, so Dan Castellaneta did the voice." It's a good impression and people didn't seem very upset about that. I am not sure how this is different. (Apparently this particular "impression" predates the Her character, so it's even easier to not be mad about. It's just a coincidence. They weren't…

I had similar thoughts based on a podcast I listened to once about voice actors hired for film spin off merchandise and whatnot. It's very common to look for voices that approximate a fictional character's voice, that was originally done by a different actor or actress.

Thinking about that episode, I imagine the legal risk is less in trying to sound like Scarlett Johansson, and more in trying to sound like Samantha, the AI character in Her. Warner Brothers or Spike Jonze probably has some legal rights to the character, and an argument could be made that OpenAI was infringing on that. The viability of that argument probably depends on how much people conflate the two or believe that the one was meant to represent the other.

Re: OpenAI didn’t copy Scarlett Johansson’s voice for ChatGPT, records show

#37
post #9

It seems increasingly difficult for common people to protect their voices, especially when even Scarlett Johansson can't manage it. As a part-time voice actor with a unique voice, I'm concerned about what I should do if my voice is used without permission and the company denies it. How can I protect myself in such a situation?

I think we'll find voices not that unique. A voice can be zero shot encoded to a few hundred kb vector. Timbre, prosody, lots of characteristics. That's less information than a fingerprint. And more importantly, that's something you can dial in with a few knobs by simply listening by ear. It's why your brain can easily hear things in other people's voices. They're not hard signals to reproduce. Some people with flexi…

I agree on the technical aspect.

But still there are some voices that are just highly associated with just one person in everyone's minds, like David Attenborough. For example, if I heard Attenborough speaking my local train announcements, but it would be an impersonator, I think I would feel like the company is taking advantage of Attenborough's voice. I.e. they would be using the fact that everyone knows this voice to their advantage, without actually paying Attenborough.

While voices aren't technically that unique, when linked to certain situations or when heard by enough people, they become unique in that context. I'm sure no one cares about Attenborough's voice 100 years from now.

Or hm, maybe AI voice tools will keep his voice alive forever in Planet Earth spinoffs, just like Sinatra has been resurrected for mashups.

Re: OpenAI didn’t copy Scarlett Johansson’s voice for ChatGPT, records show

#38
post #13

I was perusing some Simpsons clips this afternoon and came across a story to the effect of "So and so didn't want to play himself, so Dan Castellaneta did the voice." It's a good impression and people didn't seem very upset about that. I am not sure how this is different. (Apparently this particular "impression" predates the Her character, so it's even easier to not be mad about. It's just a coincidence. They weren't…

Surely there’s some kind of difference between “voice impression for a two-line cameo in one episode of an animated sitcom” and “reproducing your voice as the primary interface for a machine that could be used by billions of people and is worth hundreds of billions of dollars.” Is there a name for this AI fallacy? The one where programmers make an inductive leap like, for example, if a human can read one book to lear…

If famous actors could sue over the use of a less-famous actor that sounds just like them, what's to stop less-famous actors from suing over the use of a famous actor who sounds just like them in big-budget movies? ... and that's when you discover that "unique voice" is a one-in-a-million thing and thousands of people have the same voice, all asking for their payout.

Re: OpenAI didn’t copy Scarlett Johansson’s voice for ChatGPT, records show

#39

I can't help but think that this was all planned. It is a very intricately planned, and geniously executed marketing ploy to make sure everyone knows about the company, the new release, that there is voice now, and even makes everyone look into it just to "see for themselves". Whether this was in with ScarJo in the loop or not, does not really change the outcome, but would be a nice information we probably will never…

While I cannot say you're right or wrong, we both share similar thought! So much so that I feel like this is not the first time OpenAI has done this level of antics just to get more exposure. Seriously... I have spent pondering the same every time they get into news on the basis of drama.

It's either our delusion (you and me) or they have someone in the marketing department who has a really good grasp of how to ride the news cycle wave.

Re: OpenAI didn’t copy Scarlett Johansson’s voice for ChatGPT, records show

#40
post #18

The thing that worried me initially was that: - the original report by Scarlett said she was approached months ago, and then two days prior to launch of GPT-4o she was approached again Because of the above, my immediate assumption was that OpenAI definitely did her dirty. But this report from WaPo debunks at least some of it, because the records they have seen show that the voice actor was contacted months in advance…

I don't know, to me, it's just sounds like they know how to cover all their bases. To me, it sounds like they had the idea to make their AI sound like "her". For the initial version, they had a voice actor that sounds like the movie, as a proof of concept. They still liked it, so it was time to contact the real star. In the end, it's not just the voice, it would have been the brand, just imagine the buzz they would h…

Agreed — seems like they had a plan, and probably talked extensively with Legal about how to develop and execute the plan to give themselves plausible deniability. The tweet was inadvisable, and undoubtedly not part of the actual plan (unless it was to get PR).
Post reply on HN