Meta says its new speech-generating AI model is too dangerous for public release.
Meta announced a new AI model called Voicebox yesterday, one it says is the most versatile yet for speech generation, but it’s not releasing it yet:
There are many exciting use cases for generative speech models, but because of the potential risks of misuse, we are not making the Voicebox model or code publicly available at this time.
The model is still only a research project, but Meta says can generate speech in six languages from samples as short as two seconds and could be used for “natural, authentic” translation in the future, among other things.
i think you’re close, but having seen how facebook devs are with handling AI firsthand when I was in university, it’s probably more like:
“we were screwing around with something we didn’t understand and made something more powerful than we could control, so we’ve had to shut it down for now until we figure out how to limit what it can do so we don’t set loose the universe’s Nazi-est chatbot. also, we trained it on FB user data, so you know it’s really bad.”
No, it’s more like “We need to pump up the share price.”. An AI that can make realistic voices from quick samples isn’t anything new, there are already companies out there that can do this.