What did you say? Lyrebird will mimic a voice for you.

[WMGroomAK]

As if we didn't have enough reasons to be paranoid and security conscious already, you soon may not be able to trust the voice you hear...  Lyrebird, a Montreal bases startup, has created a voice imitation algorithm that will supposedly mimic a persons speech and shift the emotional cadence of the speaker based off of just a 1 minute audio clip of the person.

https://techcrunch.com/2017/04/25/lyrebird-is-a-voice-mimic-for-the-fake-news-era/

 

Quote

Lyrebird says its intention is to offer an API in the future so that third parties can make use of the audio mimicry technology for their own ends. So if you think fake news online is bad now, wait until there’s a tech that lets anyone generate a ‘recording’ of a person apparently incriminating themselves, trivially easily.

 

The startup does have an ethics statement on its website to confront head on what it describes as the “important societal issues” thrown up by technology’s ability to fabricate recorded evidence — in which it states:

Quote

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.

Asked if they have any concerns about putting the tech into the wild, Alexandre de Brébisson, one of the PhD students developing the deep learning tech, told TechCrunch: “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.

 

So now I guess I can't really trust whether the voice on the other end of a phone call is real or not...  I mean, this is really neat technology, but it is going to be way too easy for someone to maliciously abuse.  Even if the current technology can be filtered out via some kind of identification algorithm, as more data is collected using this, I can only imagine it getting better.  I can also think that the various international intelligence agencies will be all to glad to get their hands on something like this.

[Zando_]

All I think when I hear stuff like this is "Do they ever stop and think, 'Maybe we shouldn't invent this?'"

[HunterSkater429]

cool AF

[NumLock21]

The real thing is 1,000,000x more awesome!

 

 

[WereCat]

2 minutes ago, NumLock21 said:

The real thing is 1,000,000x more awesome!

 

 

 

What the hell, police car sound, chainsaw sound.....  That bird would be a total troll if people were teaching him sounds on purpose and let him "sing" them in the city.

[NoxiousOdor]

13 minutes ago, Zando Bob said:

All I think when I hear stuff like this is "Do they ever stop and think, 'Maybe we shouldn't invent this?'"

[HeathCliff]

48 minutes ago, NumLock21 said:

The real thing is 1,000,000x more awesome!

 

 

Oh god, the things I'd teach one of those....

[Trixanity]

Isn't it possible to embed a watermark in the recording indicating that it was forged? And couldn't an analysis of the recording reveal the forgery to begin with?

 

I mean one thing is that the human ear can't determine the difference but I figure advanced hardware and software would be able to decipher it especially if a watermark was added to it.

[mynameisjuan]

This is on the level with "virtual person" where they could manipulate video to or pictures to make it look legit. Now voice to top it all off and this will be used for nothing but causing havok

[Dabombinable]

Still doesn't beat the Lyrebird capable of mimicking sounds from Star Wars.

[suicidalfranco]

Customer support just got cheaper

[BuckGup]

12 minutes ago, suicidalfranco said:

Customer support just got cheaper

Funny thing is this has existed for almost 3 years now. The "people" that answer you on customer support lines are sometimes not always people

[tlink]

we really should stop seeing vocal confirmation as a binding contract over here, its what telephone marketing relies on and they are already shady as fuuuck. with tools like this becoming more accessible they can basically manipulate anyone who sounds vulnerable enough but wont say yes into paying.

[Thony]

With this u can do anything you want. The world most powerful propagnda.... 

 

 

[Soonercoop]

[Beskamir]

This sounds like it'll be great for text to speech. Heck maybe we can finally have AI based voice acting in games opening the doors for much more dialog.