121–130 of 174 posts
> There was an error and I still haven't implemented retry to make it invisible. You can absolutely submit your request again a few times; this is a self-healing Kubernetes cluster. Some models (voices) get more load than others and/or are scaled to fewer or more pods. There's also a rate limiter, but there aren't error messages yet.
Earlier quoted context omitted.
what are the legal implications surrounding this now ?
I don't think that there are any. Simply writing a program that outputs sounds that resemble someone's voice can't be illegal, right? You can use it in illegal ways, sure, but I don't think there are any laws concerning software like this.
Earlier quoted context omitted.
I don't think that there are any. Simply writing a program that outputs sounds that resemble someone's voice can't be illegal, right? You can use it in illegal ways, sure, but I don't think there are any laws concerning software like this.
Some of these people get paid for voiceover work, so this could get interesting
Earlier quoted context omitted.
There are half a dozen startups in this space that provide the tech. They use embedded style tokens or sliders to change the emotion, pitch, timbre, etc. I don't have links off hand, but they're not too difficult to find. These companies tend to focus on off-the-shelf turnkey solutions, so they'll have a suite of a few voice actors to choose from for different character archetypes.
Yep, that's what we're doing at https://replicastudios.com - though we have strict ethical guidelines to not clone a person's voice without their permission.
Earlier quoted context omitted.
Some of these people get paid for voiceover work, so this could get interesting
yup, I would stay as far as possible from using this wit explicit permission, just putting it out there, the project has created legal risk IMO.
Tangentially related, if you have your voice print as a security mechanism at a financial institution (Vanguard), you should ask them to turn that off.
It doesn't like spelling mistakes. Ask it to say " The Aristoocrats!" with Gilbert Goddfried.
It's currently sourcing phonemes from a lookup table called CMUdict, which is constructed by Carnegie Mellon [1]. That database has 140,000 entries, but even so, you'd be surprised how many common words are omitted. And of course it is missing terms for things like "pokemon" and "fortnite", which I had to add myself. I don't have generic grapheme -> phoneme/polyphone prediction, but that's something I look to add soo…
Earlier quoted context omitted.
I wonder if at some point we need to create legal requirements that all deepfakes have some kind of human-invisible (or visible?) fingerprint for identification, restriction on frequency range, etc. We have crypto export restrictions, why not put handcuffs on this as well which has the potential for probably larger scale harm?
Actually there might be a precedent with the Carrie Fisher from Star Wars. I believe they did use some form of virtualization after she passed..I don't know what the outcome was legally, but it definitely is in this realm.
Earlier quoted context omitted.
Such a system will always suffer from false positives and false negatives. On a more positive note, when deepfakes become a problem, we will see the emergence of a culture where unsigned authoritative content is not paid any attention.
> we will see the emergence of a culture where unsigned authoritative content is not paid any attention If current events are any indication, that culture will only emerge 30 years after the tech becomes widely usable, and in the interim will lead to absolute chaos in the form of weaponized disinformation.
Earlier quoted context omitted.
what are the legal implications surrounding this now ?
I don't think that there are any. Simply writing a program that outputs sounds that resemble someone's voice can't be illegal, right? You can use it in illegal ways, sure, but I don't think there are any laws concerning software like this.
As I said, they lost on copyright claims but these cases involved their natural voice. If the voice was a made up voice, such as a character an animated character, I wouldn't immediately dismiss the idea that it might be copyrightable.
Is this ethical?