The voice actor isn’t a Johansson imitator, and the voice isn’t an imitation.
The only similarity between the Sky voice and Johansson’s is that it’s a white American female, so by your logic a significant percentage of the US population has a case.
At any rate, Altman made clear allusions to hint that they are capable of synthesizing ScarJo's voice as a product feature. The actress retaliated saying she verbally did not consent, and now OpenAI's defense is that they hired a different actress anyway.
...which means they lied to everyone else on the capabilities of the tech, which is y'know, even worse
My understanding was that the voices sound quite similar. I haven't heard the original Sky voice so don't know. Are there any samples online?
What you can't do is USE that voice in a way that seeks to mislead (by however much) people into believing it is someone else.
I'm really not sure why people can't understand that it is intent that matters.
The company would also be liable if they used your voice and claimed it was someone more famous.
Ultimately you’re not liable for having a similar voice because you’re not trying to fool people you’re someone else. It’s the company that hired you who’s doing that.
This is why tribute acts and impressionists are fine…as long as they are clear they’re not the original artist
They could have used any accent, any tone, anyone. Literally anyone else. And it would have been fine. But they obviously copied the movie even if they used a different actress.
https://www.reddit.com/r/ChatGPT/comments/1cwy6wz/vocal_comp...
And here is voice of another actress ( Rashida Jones ):
https://www.youtube.com/watch?v=385414AVZcA
If you click through and listen please reply and answer these questions: which actress do you think is similar to the openAI sky voice? And what does that tell you about likely court result for Johansson? And having reached this conclusion yourself would you now think the other actress Rashida Jones is entitled to compensation based on this similarly test?
People are so weird on this. OpenAI screwed up, they know it, their actions show it, there isn't much to discuss here.
https://www.reddit.com/r/ChatGPT/comments/1cwy6wz/vocal_comp...
And here is voice of another actress ( Rashida Jones ):
https://www.youtube.com/watch?v=385414AVZcA
which actress do you think is similar to the openAI sky voice?
Again, given the vast range of voices (even female voices) available, choosing one that sounds so close to Her, given the subject matter of the film (and the OpenAI leadership's references to the film), this is not coincidence.
Her statement says otherwise:
"Last September, I received an offer from Sam Altman, who wanted to hire me to voice the current ChatGPT 4.0 system. He told me that he felt that by my voicing the system, I could bridge the gap between tech companies and creatives and help consumers to feel comfortable with the seismic shift concerning humans and Al. He said he felt that my voice would be comforting to people.
If the records show that they did train Sky with Johansson’s voice samples it will be an interesting case.
Here's a video that someone posted in October talking to the same Sky voice: https://www.youtube.com/watch?v=SamGnUqaOfU
This will settle out of court.
Put side by side...I can hear a similarity, but they're also distinctly different (the OAI voice doesn't have the huskiness of the Her VA IMO, but they're both Californian female with a bubbly demeanour).
And people have been making computers sound like humans without anyone suggesting that it's some attempt at fraud for very long.
The voice doesn't need to sound the same for OpenAI to lose.
> Why reference the film Her?
Because they developed an AI some people is bonding with. Which is the bigger deal? The voice or the AI, you tell me.
I'm afraid the whole world will get regulated like EU someday, crippling innovation to a point that everyone's afraid to break a law that they aren't even aware of, and stop innovating.
Er, that is totally possible? You act like it's not a machine learning system. You train new stuff in hours or days easily, especially if you have good tooling. Imagine saying this of, say, Stable Diffusion image LoRAs: "this X artist LoRA couldn't be based on X because it was somehow trained within the few weeks after X said no!"
All the timing means is that, in good management & startup fashion, because they needed multiple voices, they had a voice pipeline going, and so could have a bunch of female voices in the pipeline for optionality. And if licensing Johansson didn't work out, you have a plan B (and C, and preferably D). This is big business, you don't do things serially or not have backups: "'hope' is not a plan".
In what way?
That in no way contradicts the fact that the Sky voice was created first, although it does seem to suggest a misunderstanding by Johansson that this was to be an exclusive deal to be "the" voice, leading to the incorrect conclusion that the Sky voice was created after she declined, and must therefore be an impersonation (despite sounding nothing like her/Her, as she herself must know better than anyone). Stretch after stretch after stretch. (Being kind.)
In fact the recordings used for training were made in June/July 2023, which is before Johansson was contact as a possible "also-ran": https://openai.com/index/how-the-voices-for-chatgpt-were-cho...
Then don't worry about it. It doesn't matter anyway. Ponder the questions in the comment you replied to in stead; the ones you evaded by asking these irrelevant ones.