Our now Alumni Brett Schultz is here to impart some knowledge to us about the powerful bird, called a Lyrebird and how it’s changing what we know about recording sounds.

A Montreal-based startup called Lyrebird is a new, controversial Artificial Intelligence (AI) platform that allows users to create their own digital copy of their voice.  Or better put, anyone with audio of your voice can steal your voice.  While users should only use their own voice, Lyrebird has gained negative attention because of its misuse, particularly in fake news and false confessions.

What is it?

Like the sound of imitations of a lyrebird in nature, so are the fake voices that are used on parody videos.  Soon enough, the fake voices won’t just be for fun parody videos.  As the firm improves its technology the clarity of the audio that could be used to provide false confessions, fake news, or even audio-video overlay will become more popular.  You might have already heard audio that was from using Lyrebird, such as the voice of Barack Obama and other celebrities.

The AI is so powerful that only 60 seconds of audio is needed for the computers to pick out the pronunciation, emphasis, and characteristics of one’s voice to be able to speak words that were not included in the audio recording.  The more audio, the higher quality the voice that can be produced.  The software is available for anyone to use with a 60 second version being available for users in its BETA program.

The issues

As an IT person, I can’t help but think of the security concerns that software this powerful poses for all people.  It raises a lot of questions relating to privacy, false confessions in court, and copyright issues that are unregulated because these issues have never been seen before.  Recently, Lyrebird has been used in fake news and has sent crowds into panics about going to war, or pranks in parking lots from the emergency radio faking the emergency broadcast system that a nuclear missile is heading our way.

A limited liability Lyrebird

Lyrebird is aware of the dangerous abuse it offers, and so it offers free audio analysis directly on their website that allows people to upload audio and determine if it is fake or real.  But, is this enough to ensure that it won’t be misused?  As Lyrebird evolves in its BETA version and larger releases are available to the public, will there be a hidden signature to the audio files such as a high-pitched copyright notice that is inaudible to human ears but analysis can reveal?

Lyrebird’s website even features an ethics statement that describes the “important societal issues” being created with the likes of this type of technology.  It states:

Voice recordings are currently considered as strong pieces of evidence in our societies and in particular in jurisdictions of many countries. Our technology questions the validity of such evidence as it allows to easily manipulate audio recordings. This could potentially have dangerous consequences such as misleading diplomats, fraud and more generally any other problem caused by stealing the identity of someone else.

By releasing our technology publicly and making it available to anyone, we want to ensure that there will be no such risks. We hope that everyone will soon be aware that such technology exists and that copying the voice of someone else is possible. More generally, we want to raise attention about the lack of evidence that audio recordings may represent in the near future.

Alexandre de Brebisson, a PhD student working on the project, said ,“By releasing the API publicly and allowing anyone to use it, we want people to become aware that this technology exists and that audio recordings are not as reliable as we may think.  It is similar to what Photoshop did.  Not publishing the technology because of those potential missuses do not make sense to us as we think that the positive aspects overcome the bad ones (a hammer can be used to build but also to break).  If we do not publish the technology ourselves, others will do it in the future (and, contrary to us, they might have bad intentions, maybe hiding it from a part of the population).”

The first samples with limited features such as emotional emphasis will be free, but Lyrebird plans to implement a monetization plan for developers and companies so that they pay for the number of samples they request.

The Uses

From audio book readings with famous voices, personal assistants, speech synthesis for people with disabilities, animation, video games, and connecting devices – the uses are out there.  With technology this advanced, the full scope of its uses has not been realized, but with it, the full scope of its negative uses is also not realized, and could pose a threat to consumers.  There is very little we do that isn’t on camera, on our cell phones, or being monitored.  Hopefully, legislators will quickly address this and many other technological issues, such as taxing cryptocurrencies, sooner rather than later.

If you have any questions about Lyrebird or any tech troubles, be sure to email fsj@racc.edu with a mention of your concern in the subject.

Categories: General, Technology